Modelos de IA

Claude Haiku 4.5

Anthropic

Recomendado

A referência de custo-benefício da Anthropic a $1/$5 por 1M de tokens — o nível rápido e barato contra o qual os concorrentes se comparam, com contexto de 200K e saída de 64K. Melhor para trabalho de alto volume e sensível à latência: classificação, extração, sumarização, roteamento. Recorra ao Opus 4.8 ou ao Sonnet 5 quando for necessário raciocínio profundo em múltiplas etapas.

Entrada

$1 / 1M tokens

Saída

$5 / 1M tokens

Contexto

200K ctx; legacy, Haiku 5.5 is current

Claude Opus 4.8

Anthropic

Recomendado

Claude Opus 4.8 is Anthropic's previous-generation Opus, now listed as legacy: 1M-token context, 128K output, $5/$25 per 1M tokens, retirement not before 28 May 2027. It remains our pick for review-heavy work such as code review and constrained refactoring, where deliberate pacing beats persistence. New workloads should start on Claude Opus 5.5 at $4/$20, which Anthropic now recommends.

Entrada

$5 / 1M tokens

Saída

$25 / 1M tokens

Contexto

1M tokens

Claude Opus 5

Anthropic

Recomendado

Claude Opus 5 delivers near-Fable 5 intelligence at $5/$25 per 1M tokens, with a 1M-token context and 128K output. Anthropic now lists it as legacy: still available, retirement not before 24 July 2027, and Claude Opus 5.5 ($4/$20) is its recommended starting point. Recommended for teams already running Opus 5; new workloads should start on Opus 5.5, which is cheaper per token.

Entrada

$5.00 / 1M tokens

Saída

$25.00 / 1M tokens

Contexto

1M ctx; 128K out; legacy; fast $10/$50

Claude Sonnet 5

Anthropic

Recomendado

Anthropic's most agentic Sonnet at launch delivers performance close to Opus 4.8 at 40 percent of the Opus rate. The $2/$10 per 1M token price is the standard rate: Anthropic cancelled the increase to $3/$15 that was scheduled for 1 September 2026. The newer tokenizer produces about 30 percent more tokens for the same text, so recount budgets before migrating from Sonnet 4.6. Anthropic now lists Sonnet 5 as legacy; Claude Sonnet 5.5, at the same $2/$10, is the current Sonnet, so start new work there.

Entrada

$2 / 1M

Saída

$10 / 1M

Contexto

1M tokens; batch $1/$5; cache read $0.20

DeepSeek V4 Flash

DeepSeek

Recomendado

DeepSeek retired V4 Flash from its API on 10 September 2026: the deepseek-v4-flash id is still accepted, but its requests are served by V4.1-Flash and billed at the Flash rate of $0.15 input and $0.60 output per 1M tokens off-peak. The MIT-licensed weights (284B parameters, 13B active, 1M-token context) remain on Hugging Face, so this is now a frozen checkpoint for self-hosting and reproducible evals. For new API work, call deepseek-flash.

Entrada

$0.15 / 1M tokens

Saída

$0.60 / 1M tokens

Contexto

Retired Sep 10; id serves V4.1-Flash. Peak 2x M-F

DeepSeek V4 Pro

DeepSeek

Recomendado

Strong price-performance on open MIT weights; the 2x peak window is 01:00-04:00 and 06:00-10:00 UTC on weekdays only, so weekends bill entirely off-peak.

Entrada

$0.66 / 1M tokens

Saída

$1.98 / 1M tokens

Contexto

1M tokens; 2x price at weekday peak hours

Gemini 2.5 Pro

Google

Recomendado

Agora a escolha econômica para contexto longo: um corte de 40% no preço (junho de 2026) tornou as análises de corpus completo acessíveis para orçamentos médios. Oferece a maior janela de contexto disponível, com forte integração ao ecossistema Google.

Entrada

$1.25 / 1M tokens

Saída

$10 / 1M tokens

Contexto

1M tokens

Gemini 3.5 Flash

Google

Recomendado

O destaque de eficiência de custos do Google para automação agêntica de desktop a $1,50/$9 por 1M de tokens com contexto de ~1M. O Computer Use nativo atinge 78,4 no OSWorld-Verified — a apenas 0,3 pontos do GPT-5.5 por aproximadamente um terço do custo. A ressalva relevante: a ferramenta Computer Use é uma prévia pública que ainda não está na API de produção, portanto verifique a disponibilidade do endpoint antes de integrá-la.

Entrada

$1.50 / 1M tokens

Saída

$9 / 1M tokens

Contexto

1M tokens

Gemini 3.5 Flash-Lite

Google

Recomendado

O Gemini 3.5 Flash-Lite é o modelo mais rápido e barato da série 3.5 do Google, a 350 tok/s e $0,30/$2,50. O salto geracional em relação ao 3.1 Flash-Lite é a verdadeira história — ele supera dramaticamente seu antecessor em benchmarks agênticos chave e vence o Gemini 3 Flash completo em tarefas de codificação e uso de computador. Para desenvolvedores escalando subagentes agênticos de alta vazão, pipelines de busca e processamento de documentos, a relação preço-desempenho do Flash-Lite é inigualável na linha do Google.

Entrada

$0.30 / 1M

Saída

$2.50 / 1M

Contexto

1M context, 64K output, 350 tok/s

Gemini 3.6 Flash

Google

Recomendado

Gemini 3.6 Flash is Google's new default Flash-tier workhorse: better than 3.5 Flash on every benchmark and cheaper, at $0.75/$3.75 per 1M tokens through 2026-12-31, rising to $1.50/$7.50 from 2027-01-01. The 17% token-efficiency gain lowers per-task cost, and built-in Computer Use via the API makes it a credible agentic coding option. For Google-ecosystem developers it is the clear upgrade; budget for the January rate step.

Entrada

$0.75 / 1M (through 2026-12-31)

Saída

$3.75 / 1M (through 2026-12-31)

Contexto

1M context, 64K output; $1.50/$7.50 from 2027-01-01

Gemini 3.7 Flash

Google

Recomendado

O Gemini 3.7 Flash é o cavalo de batalha Flash mais forte do Google até agora para código e agentes — um avanço genuíno em relação ao 3.6 Flash pela metade do preço anterior até 2026. Ele ainda fica atrás da fronteira nas tarefas mais difíceis de terminal e uso de computador. Um upgrade claro para construtores de agentes sensíveis a custo e do ecossistema Google; não é um substituto de código de fronteira.

Entrada

$0.75 / 1M

Saída

$3.75 / 1M

Contexto

1M context, 64K output

Gemini 3.8 Flash

Google

Recomendado

Gemini 3.8 Flash is Google's strongest Flash workhorse yet — a genuine capability upgrade over the recommended 3.7 Flash at the same introductory price, with real gains across software engineering, agents and professional domains. The deciding caveat is Google's own: 3.8 "works harder," so per-task token burn runs well above 3.7, which remains the efficiency-first pick. Adopt 3.8 for long-horizon coding, agents and finance or legal work — benchmark your own workload before the intro price expires.

Entrada

$0.75 / 1M

Saída

$3.75 / 1M

Contexto

1M context, 64K output

Gemini 3.8 Live Extended Thinking

Google

Recomendado

Gemini 3.8 Live Extended Thinking is the voice-agent model to start with: it tops Artificial Analysis' Speech to Speech Index and reasons and calls tools in the background while it keeps talking, at the same rate as the default Gemini 3.8 Live. The caveat is the Live API surface: no structured outputs, caching, code execution or batch. Right for teams building real-time voice agents that must solve multi-step problems mid-call.

Entrada

$0.75 / 1M (text)

Saída

$4.50 / 1M (text)

Contexto

131K in, 65K out; audio $3/$12 per 1M

GPT-4.1

OpenAI

Recomendado

O destino padrão de migração para cargas de trabalho do GPT-4 Turbo: mesmo formato de requisição, custo menor, contexto de 1M tokens — converta payloads legados de function_call para o formato tools e siga em frente.

Entrada

$2 / 1M tokens

Saída

$8 / 1M tokens

Contexto

1M tokens

GPT-5.5

OpenAI

Recomendado

GPT-5.5 is OpenAI's April 2026 general-purpose model at $5 input and $30 output per 1M tokens, with a 1M+ token context (922K in, 128K out). It is no longer OpenAI's only deployable frontier option: the GPT-5.6 family is on the public price list, and GPT-6 Sol lists at $2/$10. It stays recommended as a stable, widely supported production baseline, but new workloads should test GPT-6 Sol or GPT-5.6 Terra first.

Entrada

$5 / 1M tokens

Saída

$30 / 1M tokens

Contexto

1M+ tokens (922K in / 128K out)

GPT-5.6 Sol

OpenAI

Recomendado

GPT-5.6 Sol is no longer OpenAI's best-value frontier model: its successor GPT-6 Sol, released 22 September 2026, lists at $2/$10 per 1M tokens against Sol's $4/$20. It stays recommended because its coding lead has third-party backing, while every GPT-6 Sol score so far is OpenAI's own. OpenAI lists no deprecation, and the $4/$20 rate is promotional through at least 21 November 2026. Test GPT-6 Sol before committing new workloads.

Entrada

$4 / 1M tokens

Saída

$20 / 1M tokens

Contexto

Short $4/$20 · long $8/$30 per 1M · promo

GPT-Live-Transcribe

OpenAI

Recomendado

O modelo de voz para texto com streaming de baixa latência da OpenAI entrega deltas de transcrição ao vivo a US$ 0,017/min com latência ajustável e transcrição sensível a contexto.

Entrada

$0.017 / minute

Saída

N/A

Contexto

Per-minute realtime audio billing

GPT-Transcribe

OpenAI

Recomendado

O mais recente modelo de transcrição de fala em lote da OpenAI pontua 3,31% AA-WER a $0,0045/min — 25% mais barato e mensuravelmente mais preciso que seu antecessor gpt-4o-transcribe. Suporta prompts de contexto, dicas de palavras-chave e detecção de idioma em mais de 22 idiomas, tornando-o a melhor opção padrão para transcrição assíncrona de arquivos na plataforma OpenAI. Ressalva: sem diarização ou timestamps em nível de palavra; use gpt-4o-transcribe-diarize para rótulos de falante ou gpt-live-transcribe para necessidades em tempo real. Melhor para desenvolvedores criando transcrição de reuniões, processamento de mídia e interfaces de voz na API da OpenAI.

Entrada

$0.0045 / minute

Saída

N/A

Contexto

Per-minute audio billing

Claude Fable 5

Anthropic

Condicional

Anthropic's Mythos-class frontier model, the strongest publicly available Claude at its June 2026 launch, now listed as legacy behind Claude Fable 5.1. The verdict stays conditional on two constraints: API pricing of $10/$50 per 1M tokens, and a safety classifier that over-triggered on routine coding tasks at launch, routing them to an Opus 4.8 fallback. Subscribers reach Fable through Claude Max at 50% of weekly limits, or through usage credits on Pro. Anthropic's current lineup lists Fable 5.1 and Opus 5.5 for new work.

Entrada

$10 / 1M tokens

Saída

$50 / 1M tokens

Contexto

1M ctx, 128K out; legacy; Max plans at 50% limits

Claude Fable 5.1

Anthropic

Condicional

Anthropic's new frontier flagship is the most capable model publicly available, and it largely fixes the safeguard over-triggering that made Fable 5 frustrating for routine coding. The verdict stays conditional on cost: independent measurement shows more spend per task than Fable 5 despite the cheaper cache reads, because the model thinks longer. Reach for it on demanding reasoning and long-horizon agentic work. For everyday coding, Claude Opus 5.5 at $4/$20 per 1M is Anthropic's recommended starting model and the better default.

Entrada

$10 / 1M tokens

Saída

$50 / 1M tokens

Contexto

1M context, 128K output; cache $0.25/1M

Claude Haiku 5.5

Anthropic

Condicional

Claude Haiku 5.5 is a conditional pick: Anthropic's fastest and cheapest current Claude at $0.10 input and $0.50 output per 1M tokens, with a 1M context and adjustable effort. Prompts over 100K tokens cost five times as much, and every benchmark is Anthropic's own. Right for high-volume classification, extraction, routing and subagent work. Use Sonnet 5.5 for complex agentic coding, where Anthropic's own Terminal-Bench 4.0 numbers put it well ahead.

Entrada

$0.10 / 1M tokens

Saída

$0.50 / 1M tokens

Contexto

1M ctx; prompts over 100K cost $0.50/$2.50

Claude Mythos 5.1

Anthropic

Condicional

Same weights as Claude Fable 5.1 at the same $10/$50 rate, with cyber and life-sciences safeguards loosened for vetted users. The catch is access: Mythos 5.1 ships only through Anthropic's trusted-access programs (Cyber Verification and Life Sciences Verification), so there is no self-serve API and no way to buy your way in. Conditional for the same reason as Mythos 5: elite capability behind a vetting gate. Everyone else should use Fable 5.1.

Entrada

$10 / 1M tokens

Saída

$50 / 1M tokens

Contexto

1M context, $0.25/1M cache; vetted access

Claude Opus 5.5

Anthropic

Condicional

Claude Opus 5.5 is the model Anthropic now recommends starting with for most workloads, built for long-running agentic coding and knowledge work at rates 20% below Claude Opus 5. Conditional because every benchmark is Anthropic's own and migration is not drop-in: thinking cannot be turned off, forced tool use returns an error, and most cybersecurity tasks are rerouted to Opus 4.8. Right for Claude Opus 5 teams that can absorb those breaking changes.

Entrada

$4 / 1M tokens

Saída

$20 / 1M tokens

Contexto

1M ctx; 128K out; cache $0.20; fast $8/$40

Claude Sonnet 4.6

Anthropic

Condicional

Superseded by Claude Sonnet 5 (shipped Jun 30, 2026), which Anthropic now lists as legacy behind Claude Sonnet 5.5 as the current Sonnet. Both list at $2/$10 per 1M tokens against Sonnet 4.6's $3/$15, since Anthropic cancelled the scheduled rise to $3/$15. Sonnet 5 also adds a 13.4-point Terminal-Bench 2.1 jump. Choose 4.6 only for pinned deployments or where Sonnet 5's ~30% higher token count and removed temperature/top_p/top_k are disqualifying.

Entrada

$3 / 1M tokens

Saída

$15 / 1M tokens

Contexto

1M tokens

Claude Sonnet 5.5

Anthropic

Condicional

Conditional. Claude Sonnet 5.5 is Anthropic's default Sonnet model on the Anthropic API, on the same rate card as Sonnet 5, and Anthropic says it runs faster and costs less per task. Every benchmark is Anthropic's own, and five breaking changes hit code written for Sonnet 5. Right for teams ready to re-test agentic coding and knowledge work; hold off if you rely on forced tool choice or temperature control.

Entrada

$2 / 1M tokens

Saída

$10 / 1M tokens

Contexto

1M ctx; batch $1/$5; cache read $0.10

DeepSeek V4.1 Flash

DeepSeek

Condicional

DeepSeek V4.1 Flash, the model now behind DeepSeek's deepseek-flash API, is a strong-value open-weight pick: MIT-licensed weights, image input, a 1M-token context and an Artificial Analysis score well above average for comparable models. Conditional because its standout coding-agent results are DeepSeek's own, Artificial Analysis measured it as very verbose, and peak hours double the rate. Right for cost-sensitive pipelines that can run off-peak or self-host; wrong for buyers who need independently verified agent performance before committing.

Entrada

$0.15 / 1M tokens

Saída

$0.60 / 1M tokens

Contexto

1M tokens; 2x price at weekday peak hours

Conditional. Eleven v4 is ElevenLabs' long-form text-to-speech model (id eleven_v4), covering 90+ languages with multi-speaker dialogue and voice cloning, and accepting up to 10,000 characters per request. The deciding caveat is price: $0.022 per 1K characters is a 72%-off promo that ends Oct 12 against a $0.08 list rate, and the quality claims are ElevenLabs' own, untested by us. Suits teams producing content and long-form audio who budget at list price.

Entrada

$0.08 / 1K characters

Saída

Billed per character (single rate)

Contexto

List $0.08 / 1K chars; 10K char limit

Conditional. Eleven v4 Turbo (id eleven_v4_turbo) is ElevenLabs' real-time speech model, stating a median inference latency of about 100 ms, excluding application and network latency, across 90+ languages via a Text to Dialogue WebSocket. The deciding caveat is that the latency figure is vendor-stated and untested by us, and $0.011 per 1K characters is a promo ending Oct 12 against a $0.04 list rate. Suits builders of voice agents and interactive characters.

Entrada

$0.04 / 1K characters

Saída

Billed per character (single rate)

Contexto

List $0.04 / 1K chars; median latency ~100 ms

FLUX 3

Black Forest Labs

Condicional

FLUX 3 is Black Forest Labs' unified multimodal model for image, video, audio and robot-action prediction. Its Video component is now pay-as-you-go through BFL's dashboard and API, from $0.06 per second at Draft HD to $0.80 at UHD, with five seconds of Standard HD costing $0.85. Conditional because every preference result is BFL's own, Image and Audio are not yet released, and the open-weight FLUX 3 Dev is still listed as coming soon.

Entrada

$0.17 / sec (Standard HD text-to-video)

Saída

Included in per-second video rate

Contexto

Video GA pay-as-you-go; $0.06-$0.80/s by res

Gemini 3.1 Pro

Google

Condicional

O modelo Pro mais capaz do Google disponível publicamente, com contexto de 1M tokens a preços competitivos. Geração anterior: Gemini 3.6 e 3.5 Flash agora o superam em benchmarks de codificação e agentes com custo menor. Mantenha-o no Google Cloud/Vertex AI; caso contrário, a linha Flash ou os concorrentes entregam mais.

Entrada

$2 / 1M

Saída

$12 / 1M

Contexto

1M tokens

Gemini 3.8 Flash Cyber

Google

Condicional

Gemini 3.8 Flash Cyber is Google's most capable cybersecurity model, a defensive specialist already in production at Chrome Security, Cloud Vulnerability Research and partner Wiz, with frontier vulnerability-discovery and patching results at a fraction of the leaders' rollout cost. The deciding caveat is access: it is gated to vetted defenders through the Fairwind Program, with no public API or published rate card. We rate it conditional rather than caution like the sibling Gemini 3.5 Flash Cyber because 3.8 ships published benchmark figures and named production deployments where 3.5 had neither, even though both remain gated. Right for organizations that can qualify as trusted defenders; everyone else cannot buy it today.

Entrada

N/A (no public rate card)

Saída

N/A (no public rate card)

Contexto

Fairwind-gated; no public API

Gemini 3.8 Flash TTS

Google

Condicional

Conditional. Gemini 3.8 Flash TTS is Google's flagship text-to-speech model, available in the Gemini API with voice design, consented voice replication and SynthID watermarking on every clip. The deciding caveat is cost: today's paid rates are launch prices that double when the launch window closes, and its Hume AI leaderboard placements are Google-reported, not measured by us. Suits developers building expressive narration, character voices or dubbing who can budget for the post-launch rate.

Entrada

$0.50 / 1M (text)

Saída

$9.00 / 1M (audio)

Contexto

Launch rate to 2026-12-31; $1/$18 from 2027

Gemini 3.8 Flash-Lite TTS

Google

Condicional

Conditional. Gemini 3.8 Flash-Lite TTS is Google's cost tier for high-volume speech: stable in the Gemini API, built into Google Vids, and billed at the same text input rate as Flash TTS with a lower audio output rate. Google ranks it second only to Flash TTS on Hume AI's Overall Quality Index, but that ranking is Google-reported and its launch rates are temporary, doubling once the promotion ends. Suits voice-agent cascades and bulk narration where cost per clip matters more than acting range.

Entrada

$0.50 / 1M (text)

Saída

$6.00 / 1M (audio)

Contexto

Launch rate to 2026-12-31; $1/$12 from 2027

Gemini 3.8 Live

Google

Condicional

Gemini 3.8 Live is Google's default Live API model for low-latency voice agents that keep talking while tool calls finish in the background. Conditional because Google bills Gemini 3.8 Live Extended Thinking at the identical rate, and Artificial Analysis scores that sibling higher on its Speech to Speech Index, so you pay the same for less reasoning. Pick 3.8 Live only when turn latency matters more than reasoning depth: developers shipping real-time voice experiences on the Gemini API.

Entrada

$0.75 / 1M (text)

Saída

$4.50 / 1M (text)

Contexto

131K in, 65K out; audio $3/$12 per 1M

GLM-5.2

Z.ai

Condicional

Melhor modelo de código com pesos abertos, 1M tokens de contexto e licença MIT. Supera o GPT-5.5 em múltiplos benchmarks por uma fração do custo, mas ainda fica atrás do Claude Opus 4.8 nas tarefas mais difíceis de longa duração.

Entrada

$1.40 / 1M tokens

Saída

$4.40 / 1M tokens

Contexto

1M tokens

GLM-5.3

Z.ai

Condicional

GLM-5.3 is Z.ai's open-weights coding and cyber-defense model on the GLM-5.2 base, with every gain from post-training. Pricing is published at $1.40 input / $4.40 output per 1M, the same as GLM-5.2, and the weights are now on Hugging Face. It stays conditional because its headline coding and cyber scores come from Z.ai's own evaluations: reach for defensive security and agentic automation, but audit it yourself before standardizing.

Entrada

$1.40 / 1M tokens

Saída

$4.40 / 1M tokens

Contexto

1M ctx; Flash $0.15/$0.50, launch promo ended

GPT-4o (Retired from ChatGPT, API-only)

OpenAI

Condicional

GPT-4o (May 2024) has been superseded by the GPT-5.x family and delisted from OpenAI's flagship API pricing page. It remains available through the API but is no longer a current-generation model. Use only where migration to GPT-5.x is blocked; otherwise, GPT-5.5 or GPT-5.4 offer better performance at comparable cost.

Entrada

$2.50 / 1M tokens

Saída

$10 / 1M tokens

Contexto

128K tokens

GPT-5.5-Cyber

OpenAI

Condicional

85,6% no CyberGym — maior pontuação de modelo único — mas acesso restrito apenas para defensores credenciados.

Entrada

N/A

Saída

N/A

Contexto

Trusted Access for Cyber; not commercial API

GPT-5.6 Luna

OpenAI

Condicional

GPT-5.6 Luna is OpenAI's previous-generation budget model at $0.20 input and $1.20 output per 1M tokens short context, generally available through the API and Codex. Its successor GPT-6 Luna, released 22 September 2026, lists at $0.10/$0.50, so Luna is no longer the low-cost default for new OpenAI work. It remains conditional because benchmarks are vendor-reported and it lacks activation classifiers; OpenAI lists no deprecation, so existing pipelines need not migrate yet.

Entrada

$0.20 / 1M tokens

Saída

$1.20 / 1M tokens

Contexto

Short $0.20/$1.20 · long $0.40/$1.80 per 1M

GPT-5.6 Terra

OpenAI

Condicional

GPT-5.6 Terra is OpenAI's mid-tier workhorse: GPT-5.5-class capability on OpenAI's own claim, at $2.00 input and $12.00 output per 1M tokens at short context ($4.00 / $18.00 long) after the 20% price cut of 30 July 2026. It is generally available via the API and Codex. Remains conditional: all benchmarks are vendor-reported without independent reproduction, and Terra is not selectable in the standard ChatGPT model picker.

Entrada

$2.00 / 1M tokens

Saída

$12.00 / 1M tokens

Contexto

Short $2.00/$12.00 · long $4.00/$18.00 per 1M

GPT-5.6-Cyber

OpenAI

Condicional

O GPT-5.6-Cyber é o modelo de cibersegurança da OpenAI treinado especificamente para esse fim, construído sobre o GPT-5.6 Sol e disponível exclusivamente através do programa restrito Daybreak Red. É o primeiro modelo a atingir o limiar 'Alto' de capacidade cibernética da OpenAI, com desempenho em cadeias de exploit e zero-day que supera em muito os modelos de propósito geral.

Entrada

$12.50 / 1M tokens

Saída

$75 / 1M tokens

Contexto

400K tokens

GPT-6 Astra

OpenAI

Condicional

Astra ships at $10/$50 per 1M tokens short context, $20/$75 long context, with a 1,050,000-token context window and 128K max output. On OpenAI's own reported benchmarks it leads on computer use, terminal work, long context and cyber. We previously placed it third on the Artificial Analysis Intelligence Index at 61.2; we could not source that figure and have withdrawn it, and Artificial Analysis publishes no Astra score, so we have no vendor-neutral general-intelligence ranking for it. Conditional on two constraints: Fast mode is unavailable under EU data residency, and OpenAI disclosed decreased monitorability versus GPT-5.6 Sol. Route to it by task; do not make it your default.

Entrada

$10 / 1M tokens

Saída

$50 / 1M tokens

Contexto

1.05M tokens; short-ctx standard rate

GPT-6 Luna

OpenAI

Condicional

GPT-6 Luna is OpenAI's new budget model and the cheapest way into GPT-6, at half GPT-5.6 Luna's input rate, with a 1.05M-token context and text and image input. Conditional because every published result is OpenAI's own and prompts over 272K input tokens bill at higher rates. GPT-5.6 Luna stays on the price list with no deprecation notice, so migration is optional. Right for teams running high-volume, everyday API work at scale.

Entrada

$0.10 / 1M tokens

Saída

$0.50 / 1M tokens

Contexto

1.05M ctx; over 272K $0.20/$0.75; cache $0.01

GPT-6 Sol

OpenAI

Condicional

GPT-6 Sol is the lower-priced companion to GPT-6 Astra and succeeds GPT-5.6 Sol at half its per-token rate, with a 1.05M-token context and 128K output. Conditional because every published score is OpenAI's own and prompts over 272K input tokens bill at higher rates. GPT-5.6 Sol stays on the price list with no deprecation notice, so nothing forces a migration. Right for teams routing hard coding and agent work to OpenAI.

Entrada

$2 / 1M tokens

Saída

$10 / 1M tokens

Contexto

1.05M ctx; over 272K $4/$15; cache $0.20

GPT-6.1 Sol

OpenAI

Condicional

Conditional on evidence: GPT-6.1 Sol is OpenAI's model for complex coding, computer use and professional work, which OpenAI itself describes as near-Astra performance at a lower cost. We could not read the launch post, so no benchmark is stored and the quality claim is OpenAI's alone. Prompts past 272K input tokens bill at higher rates, and reasoning cannot be set to none or minimal. Suits teams building agents on the Responses API who will run their own evals first.

Entrada

$2 / 1M tokens

Saída

$10 / 1M tokens

Contexto

1.05M ctx; over 272K $4/$15; cache $0.10

GPT-Live-1

OpenAI

Condicional

GPT-Live-1 is OpenAI's full-duplex voice model: it listens and speaks at the same time and hands reasoning and tool use to a backend model. Since 10 September 2026 it is generally available in the API on the Live endpoint at $0.05 per minute, with backend model and tool usage billed separately. Conditional because its benchmarks are OpenAI's own, language quality is uneven, and real cost per call depends on the backend model you pair it with.

Entrada

$0.05 / minute (voice session)

Saída

Included in per-minute session rate

Contexto

Live API; backend model + tools billed separately

GPT-Live-1 mini

OpenAI

Condicional

A variante gratuita do GPT-Live-1 — mesma arquitetura duplex completo em um pacote menor e mais rápido, servindo como modelo de voz padrão para usuários do ChatGPT Free. Compartilha as inovações principais: interação contínua, suporte a wake word e delegação ao GPT-5.5. As mesmas limitações se aplicam: sem API, qualidade de idioma desigual. Uma atualização substancial em relação ao antigo modo de voz baseado em turnos para usuários gratuitos.

Entrada

Free (ChatGPT)

Saída

Free (ChatGPT)

Contexto

ChatGPT Voice only

Grok 4.5

SpaceXAI

Condicional

Competitively priced coding and agentic model at $2/$6 per 1M tokens up to 200K tokens of context, doubling to $4/$12 above that, with a 500K window. xAI's published benchmarks place it mid-pack behind Fable 5 and GPT-5.5, though it leads Opus 4.8 on DeepSWE and Terminal-Bench 2.1. The real story is cost efficiency: 4.2x more token-efficient than Opus 4.8 on SWE-Bench Pro. Best for teams needing strong coding at a fraction of frontier pricing; xAI's docs list only US regions.

Entrada

$2 / 1M tokens

Saída

$6 / 1M tokens

Contexto

500K ctx; 200K+ $4/$12; cache $0.30 (200K+ $0.60)

Grok 4.6

SpaceXAI

Condicional

Grok 4.6 is SpaceXAI's frontier model for coding and agentic work at $2 input and $6 output per 1M tokens under 200K, with a 500K-token context; prompts of 200K or more bill at $4/$12. At launch it scored 61 on the Artificial Analysis Intelligence Index, tying GPT-5.6 Sol, but that is still one independent evaluator. Grok 4.7 now lists at the same price, so compare the two before committing new work.

Entrada

$2.00 / 1M tokens

Saída

$6.00 / 1M tokens

Contexto

500K ctx; 200K+ $4/$12; cache $0.50

Grok 4.7

SpaceXAI

Condicional

Grok 4.7 keeps Grok 4.6's price and speed, and xAI says it works longer on hard tasks and checks its own work more carefully. Conditional because every published score is xAI's own, prompts of 200K tokens or more bill at double the headline rate, and the Batch API is not supported. Right for coding teams already working in Cursor or Grok Build.

Entrada

$2.00 / 1M tokens

Saída

$6.00 / 1M tokens

Contexto

500K ctx; 200K+ $4/$12; fast variant 2x price

Grok Voice 3.0

SpaceXAI

Condicional

Grok Voice 3.0 is SpaceXAI's speech-to-speech model for real-time voice agents, priced at $0.08 per minute, the same as Grok Voice 2.0. Artificial Analysis shows the highest Speech-to-Speech Index score, 86.9, with 0.81 seconds to first audio, but xAI's other benchmark and customer figures are its own. Conditional because we have not tested it and the grok-voice-latest alias moved to it automatically. It suits teams building phone or in-app voice agents on a managed model.

Entrada

$0.08 / min of audio (speech-to-speech)

Saída

Included in per-minute audio rate

Contexto

Voice Agent API; $4.80 / hr

Grok Voice Think Fast 2.0

SpaceXAI

Condicional

O modelo de voz speech-to-speech de última geração da SpaceXAI com uma arquitetura de raciocínio paralelo que o torna substancialmente mais inteligente que os modelos de voz convencionais.

Entrada

$0.08 / min of audio (speech-to-speech)

Saída

Included in speech-to-speech rate

Contexto

API-only

Grok Voice Transcribe 2.0

SpaceXAI

Condicional

Grok Voice Transcribe 2.0 is SpaceXAI's speech-to-text model at $0.10 per hour of audio in batch, with speaker labels, word timestamps and key terms included. Conditional because every accuracy figure is SpaceXAI's own: it says the model is twice as accurate as 1.0 and first among 32 streaming models on Artificial Analysis, and we could not check that leaderboard. Right for voice-agent and call-transcription developers who will test it on their own audio; we have not tested it.

Entrada

$0.10 / hour of audio (batch)

Saída

Included in per-hour audio rate

Contexto

Streaming $0.20 / hour of audio

Hy3

Tencent

Condicional

Apache 2.0 open-weight MoE agentic model (295B total, 21B active) with strong tool-calling reliability and anti-hallucination behavior, but coding trails GLM-5.2 and self-hosting demands 8+ GPUs. The free OpenRouter tier that ran through Jul 21 is gone; OpenRouter now lists six providers at $0.13-$0.20 input and $0.53-$0.80 output per 1M tokens. Best for self-hosted coding agents and long-context document work where permissive licensing matters more than absolute benchmark scores.

Entrada

$0.20 / 1M

Saída

$0.80 / 1M

Contexto

256K tokens; OpenRouter providers $0.13-$0.20 in / $0.53-$0.80 out; free tier ended

Inkling

Thinking Machines Lab

Condicional

O Inkling é a razão mais forte até agora para levar a sério a IA multimodal de pesos abertos. Ele oferece raciocínio e codificação competitivos com a melhor segurança de pesos abertos da categoria e esforço de raciocínio controlável, tudo sob Apache 2.0. Mas a Thinking Machines é honesta ao admitir que não é o mais forte em termos gerais — a factualidade deixa a desejar (SimpleQA 43,9%) e a auto-hospedagem exige hardware sério (2TB VRAM). Condicional: uma base atraente para organizações que precisam possuir e ajustar um modelo multimodal, não para equipes que buscam desempenho máximo pronto para uso.

Entrada

$1.87 / 1M tokens

Saída

$4.68 / 1M tokens

Contexto

Up to 1M tokens native; Tinker 64K and 256K tiers; rates at 50% limited-time discount (list $3.74 / $9.36)

Kimi K3

Moonshot AI

Condicional

Moonshot AI's open-weight MoE model (Kimi K3 License, weights on Hugging Face) offers a 1M-token context at $3/$15 per 1M tokens and holds #1 on Arena WebDev (preliminary), but still trails Fable 5 and GPT-5.6 Sol on precise execution, and serving the weights needs 64+ accelerators. The rating stays conditional on legal risk: on July 22, 2026 the White House OSTP accused Moonshot of distilling Fable 5 and Treasury threatened sanctions. We have not re-verified that dispute's status this run.

Entrada

$3.00 / 1M

Saída

$15.00 / 1M

Contexto

1M ctx, $0.30 cached. Weights live (Jul 27).

Kolibri-1 is Aleph Alpha's English-German mixture-of-experts model with Apache 2.0 weights on Hugging Face, and a credible open-weight option for teams that must run German-language workloads on their own hardware. The catch is operations: the FP8 weights need about 78 GB of GPU memory, the pages we read name no hosted API or price, and every benchmark comes from Aleph Alpha's own harnesses. Right for public-sector and industrial teams that can self-host. Skip it for hosted-API needs or other languages.

Entrada

Free (open weights)

Saída

Free (open weights)

Contexto

262K native, 1M extrapolated; ~78 GB FP8

Laguna S 2.1

poolside

Condicional

O Laguna S 2.1 é o primeiro modelo de codificação open weight ocidental confiável em 11 meses — um MoE de 118B que ativa apenas 8B de parâmetros por token e ainda assim marca 70,2% no Terminal-Bench 2.1, superando modelos 10× maiores. Mas não está na fronteira — GPT-5.6 Sol e Kimi K3 estão muito à frente — e limitações conhecidas incluem overfitting de harness e problemas com JSON. Melhor para codificação agêntica self-hosted onde a soberania de dados importa mais do que liderança absoluta em benchmarks.

Entrada

$0.10 / 1M

Saída

$0.20 / 1M

Contexto

OpenMDW-1.1 license, 1M context

Llama 4 Maverick

Meta

Condicional

O principal modelo de código aberto para equipes que precisam de auto-hospedagem, soberania de dados ou querem evitar dependência de fornecedor de API.

Entrada

Free (self-hosted) or $0.20 / 1M tokens (hosted)

Saída

Free (self-hosted) or $0.60 / 1M tokens (hosted)

Contexto

1M tokens

MAI-Code-1-Flash

Microsoft

Condicional

MAI-Code-1-Flash was Microsoft's first in-house coding model, a sparse MoE priced at $0.75/$4.50 per 1M tokens. GitHub retired it from Copilot on 10 September 2026 and names MAI-Code-1.1-Flash, at $0.20/$1.20, as the replacement; it no longer appears in Copilot's supported-models or pricing tables. Copilot users should switch to 1.1-Flash, which is cheaper and scores higher on Terminal-Bench 2.1.

Entrada

$0.75 / 1M tokens

Saída

$4.50 / 1M tokens

Contexto

Retired from Copilot Sep 10 2026; 256K ctx

MAI-Code-1.1-Flash

Microsoft

Condicional

O MAI-Code-1.1-Flash é o modelo de código de nível small renovado da Microsoft para o Copilot — um quarto do custo e 22% melhor no Terminal-Bench 2.1 do que seu antecessor de junho, agora com visão nativa. Ele foi construído para trabalho iterativo de alto volume, e não para raciocínio de fronteira, o que o torna o modelo de código sério mais barato do Copilot. Use-o em autocompleções cotidianas, refatorações e depuração assistida por imagem; use um carro-chefe para arquitetura complexa.

Entrada

$0.20 / 1M

Saída

$1.20 / 1M

Contexto

256K context

MAI-Cyber-1-Flash

Microsoft

Condicional

Microsoft's first in-house cybersecurity model, built on the MAI-Thinking-1 family and embedded inside the MDASH multi-agent harness, handles about 90% of security workloads at roughly half the cost of prior configurations and reserves GPT-5.4 for the hardest 10%. Microsoft reports 96% on CyberGym by the any-crash metric and 86.3% by the stricter final-submission metric, both from a tuned full-system configuration rather than an independent model-versus-model test. Available only through Azure AI Foundry with vetted enterprise access; no standalone pricing or public API.

Entrada

N/A (MDASH SCU consumption)

Saída

N/A (MDASH SCU consumption)

Contexto

Azure AI Foundry vetted access

Conditional. Mercury 2.5 is Inception's diffusion LLM, and raw output speed is the reason to pick it: Artificial Analysis ranks it among the fastest models it measures, and it ships tunable reasoning, parallel tool calls and structured outputs. The deciding caveat is quality: Artificial Analysis places its intelligence just below the median, and Inception's parity claims name no benchmark. Suits latency-bound voice agents, RAG pipelines and coding subagents, not hard reasoning.

Entrada

$0.20 / 1M tokens

Saída

$0.75 / 1M tokens

Contexto

260K ctx; 80% off at launch: $0.04/$0.15

Mistral Large 4

mistral

Condicional

Conditional: Mistral Large 4 is a public preview, not a finished release, since Mistral says its reinforcement learning run is still in flight and promises weights "by the end of the month". It is a mixture-of-experts model of about 1 trillion parameters, 52B active, with a 1M context on its docs page and a $1.36 input, $4.18 output list price per 1M tokens. Most benchmarks are Mistral's own. Trial it now if you want an open-weight-bound model, but wait for weights and licence before committing.

Entrada

$1.36 / 1M tokens

Saída

$4.18 / 1M tokens

Contexto

1M ctx; preview; sale $0.68/$2.09 on docs

Muse Glimmer

Meta

Condicional

O Muse Glimmer é o modelo agentivo de pesos abertos de 30B da Meta, destilado do modelo de fronteira fechado Muse Spark e lançado sob Apache 2.0. Com quantização de 4 bits, ele executa chamadas de ferramentas multi-etapas, raciocínio multimodal e recuperação de falhas em uma única GPU de consumo de 24GB sem dependências de nuvem.

Entrada

Free (open weights)

Saída

Free (open weights)

Contexto

131K tokens

Muse Spark 1.1

Meta

Condicional

O primeiro modelo de IA pago da Meta e primeiro disparo no mercado comercial de codificação — US$ 1,25/US$ 4,25 por 1M de tokens com 1M de contexto, orquestração nativa de subagentes, suporte a MCP/ferramentas e capacidade de computer use por cerca de 4× abaixo dos preços da Anthropic/OpenAI. Lidera em benchmarks de uso de ferramentas, mas fica atrás do Opus 4.8 em codificação pura por 7 pontos no SWE-Bench Pro. Peso fechado, prévia pública apenas nos EUA. Melhor para equipes que priorizam orquestração de uso de ferramentas a baixo custo.

Entrada

$1.25 / 1M tokens

Saída

$4.25 / 1M tokens

Contexto

1M tokens

Muse Spark 1.2

Meta

Condicional

Atualização do modelo focado em codificação da Meta, co-treinado com o harness Muse Code — 82,9% no Terminal-Bench 2.1 e 59,3% no DeepSWE 1.1, segundo apenas ao Claude Opus 5 no primeiro. Preços padrão inalterados a $1,25 entrada / $4,25 saída por 1M de tokens.

Entrada

$1.25 / 1M tokens

Saída

$4.25 / 1M tokens

Contexto

1M tokens

Muse Spark 1.3

Meta

Condicional

Muse Spark 1.3 is Meta's strongest coding and long-context model yet, and an easy drop-in for anyone already on 1.2: the standard token price is unchanged while software-engineering and million-token retrieval results move sharply up. The deciding caveat is that the headline agent numbers come from a max reasoning configuration Meta has not released, while the shipping tier runs slower and costs more per task. Right for cost-sensitive coding teams inside Muse Code; wrong for anyone who needs independently reproduced benchmarks.

Entrada

$1.25 / 1M tokens

Saída

$4.25 / 1M tokens

Contexto

1M ctx; contributor tier $0.10/$0.20

Muse Spark Open

Meta

Condicional

Muse Spark Open is Meta's open-weight release of Muse Spark, an 816B mixture-of-experts model with 42B active parameters and a 1M-token context, available as downloads and from seven hosted providers. Conditional: the Muse Model License has limited commercial terms we did not read, the Meta pages we read give no hosted price, and Meta says its full evaluation report is still to come. It suits teams that want to self-host or shop between hosts, not buyers who need audited terms or independent benchmarks.

Entrada

Free (open weights)

Saída

Free (open weights)

Contexto

1M ctx; Muse Model License (limited commercial terms); no hosted price read

Muse Wonder

Meta

Condicional

Muse Wonder is Meta's most capable model on its API, but it is still a preview and we could not read a single benchmark score behind the launch claims. Its list price is 3.2x Muse Spark 1.3 on input and 4.7x on output, and Meta says the full safety report is still to come. It suits teams already on Muse Spark who want to test it on long-horizon agent work, not buyers who need independently checked results.

Entrada

$4.00 / 1M tokens

Saída

$20.00 / 1M tokens

Contexto

1M ctx; preview; cached $0.20

Mythos 5 (Claude Mythos 5)

Anthropic

Condicional

O modelo de cibersegurança mais forte do mundo (83,8% CyberGym), construído sobre os pesos do Fable 5 com os classificadores de segurança removidos. O acesso foi suspenso após a diretiva de exportação de 12 de junho; parcialmente restaurado em 26 de junho para ~100 organizações críticas dos EUA designadas (Project Glasswing). O veredicto permanece CONDITIONAL: a capacidade é de elite, mas o acesso é controlado por designação federal. Não é uma API pública: utilizável apenas se sua organização estiver na lista do Departamento de Comércio.

Entrada

$10 / 1M tokens

Saída

$50 / 1M tokens

Contexto

1M tokens (restricted access)

Ornith-1.0

DeepReinforce

Condicional

Uma família de modelos open-source para programação agentiva com licença MIT que entrega resultados comparáveis aos melhores modelos — superando o Claude Opus 4.7 tanto no Terminal-Bench 2.1 quanto no SWE-Bench Verified — mas exige self-hosting com recursos significativos de GPU e vem de um laboratório novo com histórico limitado.

Entrada

Free (MIT license)

Saída

Free (MIT license)

Contexto

256K tokens

Palmyra X6

Writer

Condicional

O Palmyra X6 é o modelo agêntico carro-chefe da Writer para fluxos de trabalho de marketing e receita — pós-treinado sobre o GLM-5.2 de pesos abertos da Z.ai a $2/$8 por 1M de tokens, uma fração da taxa dos modelos de fronteira. Todos os números de destaque são alegados pela Writer, e a proveniência do GLM-5.2 é uma questão real de procurement. Adote-o dentro da Writer para trabalho de agente GTM sensível a custo; procure em outro lugar por capacidade geral com benchmark independente.

Entrada

$2 / 1M

Saída

$8 / 1M

Contexto

1M context, 8K max output

Qwen-Image-2.1

Alibaba

Condicional

Qwen-Image-2.1 is Alibaba's open-weights model for both text-to-image generation and editing, with native transparent RGBA output and multi-image composition from up to 10 references. Conditional because the weights ship under the Qwen Research License, which allows non-commercial use only, and we found no readable benchmark scores. Right for researchers and hobbyists running image generation and editing locally through Diffusers.

Entrada

Free (non-commercial weights)

Saída

Free (non-commercial weights)

Contexto

Qwen Research License; non-commercial

Qwen3.8-27B

Alibaba

Condicional

O Qwen3.8-27B é o modelo open-weight de 27 bilhões de parâmetros da Alibaba, agora disponível sob Apache 2.0 com 262K de contexto nativo e implantação em uma única GPU. É o companheiro auto-hospedável do carro-chefe Qwen3.8-Max de 2,4T. Os benchmarks seguem sendo números da própria Alibaba, então fica em 'conditional': um modelo real e executável cujos ganhos de codificação declarados pela fornecedora ainda aguardam reprodução independente.

Entrada

Free (open weights — self-hosted)

Saída

Free (open weights — self-hosted)

Contexto

262K native ctx, 1M extended; Apache 2.0

Qwen3.8-Omni-Flash

Alibaba

Condicional

Qwen3.8-Omni-Flash takes text, images, audio and video in one call with a 1M-token context, and Alibaba says it cut the per-hour price of audio input by more than 98% against the previous version. Conditional because every benchmark we found is Alibaba's own and the model answers in text only; spoken replies need a separate model, such as the Realtime variant, at higher rates. Right for teams building audio-visual agents on Alibaba Cloud.

Entrada

$0.15 / 1M

Saída

$0.47 / 1M (text)

Contexto

1M ctx; cache hit $0.016/1M; International

Sakana Fugu

Sakana AI

Condicional

Orquestrador multiagente que iguala modelos de fronteira em benchmarks-chave ao rotear por um pool de agentes especialistas — mas o roteamento opaco e a dependência de plataforma limitam a transparência.

Entrada

$5 / 1M tokens

Saída

$30 / 1M tokens

Contexto

272K tokens (>272K: $10/$45 per 1M)

SWE-2

Cognition

Condicional

SWE-2 is Cognition's coding model for Devin, post-trained from Kimi K3, and Cognition claims it lands within one point of Fable 5.1 on FrontierCode 1.1 Main at 64% lower cost. Conditional because every SWE-2 score is vendor-reported, the same table shows it far behind Fable 5.1 and GPT-6 Astra on Terminal-Bench 4, and it runs only inside Devin. Right for Devin Desktop and CLI users; wrong for anyone who needs an API model or independently verified results.

Entrada

N/A (no standalone API)

Saída

N/A (no standalone API)

Contexto

Devin only; Enterprise $0.75 in/$3.75 out per 1M

Ternary Bonsai 2 27B

PrismML

Condicional

Ternary Bonsai 2 27B squeezes Qwen3.8 27B into about 5.9GB at 1.76 effective bits, so a 262K-context text-and-image model fits on a Mac or a single consumer GPU under Apache 2.0. Conditional because the quality figures are PrismML's own, its blog and model card report different aggregate scores, and the files need PrismML's custom kernels: stock llama.cpp will not run them. Right for local-inference users short on memory.

Entrada

Free (open weights)

Saída

Free (open weights)

Contexto

262K ctx; self-hosted; PrismML kernels required

Gemini 3.5 Flash Cyber

Google

Cuidado

O Gemini 3.5 Flash Cyber é o primeiro modelo do Google especializado em cibersegurança, com fine-tune do 3.5 Flash para encontrar, validar e corrigir vulnerabilidades. Dentro da arquitetura multiagente do CodeMender, ele atinge desempenho competitivo no CyberGym contra modelos muito maiores como o Mythos por uma fração do custo. Mas as restrições são severas: roda exclusivamente dentro do CodeMender, disponível apenas para governos e parceiros confiáveis através de um piloto de acesso limitado, sem API pública, sem números de benchmark publicados e sem cronograma para lançamento mais amplo.

Entrada

N/A

Saída

N/A

Contexto

CodeMender-only, government-gated pilot

Gemini 3.5 Pro

Google

pending

O modelo de fronteira pré-GA do Google, agora em seu quarto adiamento e 'meses atrasado' segundo a Bloomberg. O contexto de 2M tokens, o pipeline multimodal nativo e o raciocínio Deep Think continuam sendo capacidades reais, mas não há um cronograma de GA confirmado. A capacidade de codificação é a principal deficiência.

Entrada

Not yet published

Saída

Not yet published

Contexto

Est. $5-15/$30-60 per 1M. 2M ctx. Not GA.

Gemini 4

Google

pending

Modelo de próxima geração confirmado pelo Google DeepMind. Pré-treinamento iniciado em 21 de julho de 2026. Sem benchmarks, sem preços, sem especificações, sem data de GA.

Entrada

$2 / 1M tokens (through Nov 30)

Saída

$10 / 1M tokens (through Nov 30)

Contexto

$2/$10 to Nov 30, then $4/$20; paid tier only

Qwen3.8-Max

Alibaba

pending

Carro-chefe MoE multimodal de 2.4T parâmetros. Ainda não está em GA: sem preços de API por token, sem tabela de benchmarks, sem pesos abertos publicados. Entrada a $6/mês via Token Plan Lite.

Entrada

$2.00 / 1M tokens

Saída

$6.00 / 1M tokens

Contexto

Model Studio (Singapore); one tier up to 1M tokens