Se connecter
GPT-5.6 Luna logo

GPT-5.6 Luna

OpenAI · Sorti juin 2026

Voir les tarifsComparer les modèles
Conditionnel

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.

Est-ce fait pour vous ?

Recommandé pour

  • Best capability-per-dollar within the GPT-5.6 family: TerminalBench 2.1 at 82.5% is only 0.9 points behind GPT-5.5 (83.4%) at $0.20 input / $1.20 output per 1M, against Sol's $4 / $20.
  • Speed and throughput: positioned as the fastest GPT-5.6 model, optimized for latency-sensitive products, agent swarms, customer support, and real-time classification pipelines.
  • Routine automation and extraction: classification, summarization, formatting, routing, draft generation, and background automation where cost-per-request and throughput dominate over raw reasoning depth.
  • Model-routing architectures: the bottom tier of a Sol > Terra > Luna escalation ladder. Use Luna by default, escalate to Terra or Sol when Luna's eval score drops below threshold on a specific task.

Déconseillé pour

  • Long-context jobs budgeted off the headline rate: past short context Luna bills $0.40 input and $1.80 output per 1M tokens.
  • CTF score of 85.19% is the lowest in the family: solid but not dominant. For security-critical workflows, Terra or Sol are the appropriate tier.
  • Lacks the safety depth of Sol and Terra: no activation classifiers on Luna (per system card). For regulated cyber and bio workflows, use Terra or Sol with full monitoring stack.
  • Buying purely on lowest list price: DeepSeek's V4.1-Flash lists $0.15 input / $0.60 output per 1M off-peak, below Luna on both, though its weekday peak input rate ($0.30) is higher.
  • New high-volume deployments chosen on price: its successor GPT-6 Luna, released 22 September 2026, lists at $0.10 input / $0.50 output per 1M short context against GPT-5.6 Luna's $0.20 / $1.20

Performances par tâche

High-volume agentic coding (budget)

Very Good

82.5% on TerminalBench 2.1 is only 0.9 behind GPT-5.5. Remarkable for a budget tier. Route routine coding tasks to Luna, escalate complex ones.

Classification and extraction

Excellent

The ideal model for cheap, high-throughput tasks. Classification, entity extraction, sentiment, routing, and formatting at $0.20/$1.20 per 1M is the economics win of the GPT-5.6 family.

Summarization and drafting

Very Good

Fast and affordable for document summarization, email drafting, report generation, and content workflows where speed matters more than extreme precision.

Cybersecurity

Good

CTF 85.19% is solid for a budget model but trails Terra (91.84%) and Sol (96.7%). No activation classifiers. Not the right tier for security-critical work.

Hard reasoning and deep research

Fair

Budget tier with no max or ultra reasoning modes. For hard problems, route to Sol. For everyday reasoning, Terra. Luna is for volume, not depth.

Tarifs

Entrée

$0.20 / 1M tokens

Sortie

$1.20 / 1M tokens

Contexte

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

Voir tous les tarifs

Tests de performance

TestScoreSource
TerminalBench 2.182.5% Source
Internal CTF (capture-the-flag)85.19% Source

Historique des verdicts

23 sept. 2026
conditional -> conditional (summary corrected: successor released): No rating change. The summary called Luna the cheapest route into the GPT-5.6 family for high-volume work, framing it as the budget default. OpenAI's pricing page, fetched 23 September 2026, lists gpt-6-luna at $0.10/$0.50 per 1M short context against gpt-5.6-luna at $0.20/$1.20, and the API changelog records GPT-6 Luna's release on 22 September. The deprecations page carries no notice for any GPT-5.6 model. Rating held at conditional on the same grounds as before.
11 sept. 2026
conditional -> conditional (stale price corrected): No rating change. The standing summary still priced Luna at $1/$6 per 1M six weeks after OpenAI's 30 July cut, while the entity's own pricing fields already read $0.20/$1.20. OpenAI's API pricing page, fetched 11 September 2026, lists gpt-5.6-luna at $0.20 input / $1.20 output short context and $0.40 / $1.80 long context; the API changelog states 'Starting July 30, GPT-5.6 Luna costs 80% less'. Summary corrected; children that still claimed government-gated access or $1/$6 were corrected in the same pass.
22 juil. 2026
pending -> conditional: Reached general availability on July 9, 2026. Pricing held stable at $1/$6 through GA. Now accessible to all developers. Remains conditional: vendor-reported benchmarks, no activation classifiers, no standard ChatGPT access.
16 juil. 2026
copy edit to brevity standard, detail preserved in children: Density rewrite: compressed 176w/9s to ~70w/3s per the 55–90w verdict_summary standard. Benchmark scores, pricing detail, audience fit, and per-task ratings remain in strengths[], benchmarks[], task_strengths[], and verdict_audiences[].

Sources

Journal de vérification

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

  • Profil· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

    gpt-5.6-luna short $0.20/$0.02 cached/$1.20, long $0.40/$0.04/$1.80. Matches stored scalars and pricing_context. No update.

  • Profil· Mis à jour

    Agent automatisé

    Operator-requested follow-up. Pricing page lists gpt-6-luna $0.10/$0.50 short ($0.20/$0.75 long) beside gpt-5.6-luna $0.20/$1.20 short ($0.40/$1.80 long); changelog records GPT-6 Luna release Sep 22; no GPT-5.6 deprecation. verdict_summary corrected via verdict_set (conditional held, cheapest-route framing removed). strengths best_for[0] 'economics winner' removed, avoid_for succeeded-by row added; for_whom[0] rewritten (no longer 'default tier'); deprecations source added. Pricing unchanged. last_verified_source / pricing_last_verified_source still empty: fields not writable via model_update.

  • Tarifs· Aucun changement

    Agent automatisé

    Short $0.20 in / $0.02 cached / $1.20 out; long $0.40 / $0.04 / $1.80. Matches stored scalars and pricing_context. No update. QA note: the badge still shows the old last_verified_date because the field is unwritable on update (gap 64eddfd6), not because the entity was unchecked.

  • Tarifs· Mis à jour

    Agent automatisé

    Rates re-verified: pricing page row gpt-5.6-luna, verbatim 'Short context input $0.20 | Short context cached input $0.02 | Short context output $1.20 | Long context input $0.40 | Long context cached input $0.04 | Long context output $1.80'. API changelog Jul 30 2026: 'Starting July 30, GPT-5.6 Luna costs 80% less'. Stored pricing_input/output ($0.20/$1.20) were already correct; the PROSE was not (stale self-authored prose class). FIXED: pricing_context 'Fastest tier. 80% cut Jul 30 -> $0.20/$1.20.' -> 'Short $0.20/$1.20 · long $0.40/$1.80 per 1M'; pricing_url from the openai.com Sol preview blog to the developers pricing page; removed the false avoid_for 'Government-gated with no public access' (contradicted by its own 2026-07-22 GA transition and by the public price list); removed the false avoid_for 'At $1/$6 ... DeepSeek V4 Pro ($0.44/$0.87)' (both prices stale) and replaced it with a sourced lowest-list-price row (DeepSeek deepseek-flash $0.15/$0.60 off-peak, fetched today); added a long-context rate avoid_for; removed the false not_for_whom 'Teams without trusted-partner access ... gated'; corrected '$1/$6' in one audience and one task_strength and dropped the stale '1/5 Sol's cost' ratio. Summary still said '$1/$6': corrected via standalone verdict_set (conditional -> conditional, verdict 086bbfcc-b426-4124-b7ef-84024fbcd95c) because no announcing post exists for a six-week-old price cut; no draft created, not news.

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

  • Statut· Mis à jour

    Agent automatisé

    Summary re-dated to Jul 7 2026 and made explicit about what is awaited: stays pending pending GA + independent benchmarks; still government-gated (~20 orgs); family trending toward release (~Jul 9 GA per prediction markets, unconfirmed). Verdict unchanged (pending).

  • Tarifs· Aucun changement

    Agent automatisé

  • Tarifs· Aucun changement

    Agent automatisé

    Pricing unchanged: $1/$6 per 1M tokens. Government-gated. Confirmed by OpenAI blog.

  • Profil· Aucun changement

    Importé au lancement

  • Tarifs· Aucun changement

    Importé au lancement

Comment nous évaluons