neomanex
GPT-5.6 Sol logo

GPT-5.6 Sol

OpenAI · Sorti juin 2026

Voir les tarifsComparer les modèles
Recommandé

GPT-5.6 Sol est le modèle frontière le plus rentable, en tête des benchmarks de codage et compétitif sur les raisonnements difficiles pour une fraction du coût de ses pairs. Verdict renforcé par la divulgation du durcissement de sécurité GPT-Red — le modèle a été entraîné de manière adversariale contre un modèle de red teaming dédié en self-play RL à l'échelle frontière, le rendant 6x plus robuste aux injections de prompt.

Est-ce fait pour vous ?

Recommandé pour

  • TerminalBench 2.1 state-of-the-art: 88.8% base, 91.9% in Ultra subagent mode — clears Mythos 5 (88.0%) and Fable 5 (84.3%). Sol Ultra is >3 points ahead of any competitor.
  • Long-horizon coding and agentic tasks: first model past the halfway mark on Agent's Last Exam (50.9%), with max reasoning and subagent-based ultra mode for decomposing hard problems.
  • Cybersecurity and vulnerability research: matches Anthropic Mythos Preview on ExploitBench with roughly 1/3 the output tokens. CTF score of 96.7% on internal testing. Strong exploit primitive discovery in Chromium and Firefox without crossing into autonomous full-chain exploitation.
  • Biology and scientific reasoning: ~9 point overall lift over GPT-5.5 on SecureBio evaluations (Human Pathogen 68.4%, Molecular Biology 60.0%). Stronger on GeneBench v1 while using fewer tokens. Useful for legitimate biomedical research under trusted access programs.
  • Controlled release with heavy safety stack: activation classifiers, real-time misuse screening, 700K A100e GPU-hour automated red teaming, and layered model-level refusals. Heavily hardened against adversarial attacks with defensive cyber posture.
  • Prompt caching overhaul: explicit cache breakpoints, 30-minute minimum cache life, cache writes at 1.25x input rate, cache reads at 90% discount. Makes long-running agentic sessions far more predictable on cost.

Déconseillé pour

  • Long-context work costs roughly double: past the short-context tier the rate steps from $4/$20 to $8/$30 per 1M tokens. Budget by context length, not just request volume.
  • Untested on real-world messy codebases: all published benchmarks are vendor-run under vendor harnesses. The 0.8-point gap over Mythos 5 on TerminalBench is within measurement noise — these are a statistical tie.
  • At $4/$20 per 1M tokens Sol is priced for the hardest tasks. GPT-5.6 Terra delivers GPT-5.5-class performance at $2/$12, and Luna at $0.20/$1.20 — reserve Sol for problems where a weaker model's mistakes are expensive. OpenAI states current pricing is promotional at least through November 21 2026.

Performances par tâche

Codage agentique

Excellent

État de l'art sur TerminalBench 2.1 à 88,8% (91,9% Ultra). L'avance sur le peloton est réelle. La décomposition en sous-agents du mode Ultra est une avancée architecturale réelle.

Cybersécurité

Excellent

Égale Mythos Preview sur ExploitBench avec 1/3 des tokens. 96,7% CTF. Peut trouver des vulnérabilités et des primitives d'exploit. Posture défensive par conception.

Raisonnement scientifique (biologie)

Very Good

Gain d'environ 9 points sur GPT-5.5 aux évaluations SecureBio. Fort sur GeneBench v1. Human Pathogen Capabilities à 68,4%. Accès restreint pour les workflows sensibles.

Raisonnement à horizon long

Excellent

Premier modèle au-delà de 50% sur Agent's Last Exam (50,9% mode code). Le mode max reasoning et le mode ultra sous-agents poussent la profondeur de raisonnement plus loin que tout modèle antérieur.

Usage général quotidien

Good

Surdimensionné pour les tâches routinières. Terra est le meilleur choix pour le travail quotidien à moitié prix avec des capacités de classe GPT-5.5. Sol devrait être réservé aux problèmes où les erreurs d'un modèle plus faible coûtent cher.

Tarifs

Entrée

$4 / 1M tokens

Sortie

$20 / 1M tokens

Contexte

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

Voir tous les tarifs

Tests de performance

TestScoreSource
TerminalBench 2.1 (base)88.8% Source
TerminalBench 2.1 (Ultra)91.9% Source
Agent's Last Exam (code mode)50.9% Source
SecureBio: Human Pathogen Capabilities68.4% Source
SecureBio: World-Class Biology68.3% Source
SecureBio: Molecular Biology60.0% Source
SecureBio: Virology Capabilities Test53.5% Source
Internal CTF (capture-the-flag)96.7% Source

Historique des verdicts

16 juil. 2026
recommended → recommended (reinforced) — GPT-Red disclosure demonstrates frontier-scale safety investment alongside capability — the model was trained against an automated red-teamer achieving 84% attack success (6.5x human baseline) and emerged 6x more robust. This reinforces the existing Recommended stance rather than changing it.
12 juil. 2026
pending -> recommended — GPT-5.6 Sol leads the DeepSWE coding-agent benchmark at 72-73% vs Fable 5's 70% at one-third the cost, and tops the AI Analysis Coding Agent Index at 80 points. These are the independent benchmarks the directory was waiting for.
12 juil. 2026
pending -> recommended — UK AISI confirms GPT-5.6 Sol carries equivalent cyber vulnerabilities to Fable 5 — every frontier model shares this risk, which does not diminish Sol's coding and cost advantages. Vulnerability equivalence is now table stakes for frontier models.

Sources

Journal de vérification

  • Tarifs— Aucun changement

    Agent automatisé

    Hot row (promo). Pricing row verbatim: short input $4.00, cached $0.40, output $20.00; long input $8.00, cached $0.80, output $30.00. Promo note verbatim: 'GPT-5.6 Sol's promotional pricing is available at least through November 21, 2026.' Matches stored $4/$20 and pricing_context. No update, data_hash untouched. Stays hot until the promo resolves.

  • Tarifs— Aucun changement

    Agent automatisé

    Watchlist hot row (promo). Pricing page row gpt-5.6-sol, verbatim: 'Short context input $4.00 | Short context cached input $0.40 | Short context output $20.00 | Long context input $8.00 | Long context cached input $0.80 | Long context output $30.00'; promo note verbatim: 'GPT-5.6 Sol's promotional pricing is available at least through November 21, 2026.' Matches stored $4/$20, pricing_context 'Short $4/$20 · long $8/$30 per 1M · promo' and the Nov 21 strength row. No update called, data_hash untouched. Stays hot until the promo resolves.

  • Tarifs— Aucun changement

    Agent automatisé

  • Tarifs— Mis à jour

    Agent automatisé

    Pricing corrected against the live OpenAI API pricing page. Was $5/$30 per 1M sourced from the June launch post; live rate is $4 in / $20 out short-context and $8 in / $30 out long-context, described by OpenAI as promotional at least through November 21 2026. pricing_url repointed from the launch blog post to the developer pricing docs (platform.openai.com/docs/pricing now 301s to developers.openai.com/api/docs/pricing). Also removed pre-GA prose that contradicted the recommended verdict: the strengths and audiences still claimed no public access and ~20 preview organizations, while the model is listed among the generally available flagship models. Comparison figures refreshed: Terra $2/$12, Luna $0.20/$1.20.

  • 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); Sol Ultra confirmed for Codex Jul 6 by Sottiaux, prediction markets pricing Jul 9 GA (unconfirmed). Verdict unchanged (pending).

  • Tarifs— Aucun changement

    Agent automatisé

  • Tarifs— Aucun changement

    Agent automatisé

    Pricing unchanged: $5/$30 per 1M tokens. Government-gated limited preview (~20 orgs). Confirmed by OpenAI blog and aipricing.guru.

  • Tarifs— Aucun changement

    Importé au lancement

  • Profil— Aucun changement

    Importé au lancement

Comment nous évaluons