OpenAI
gpt-5.5
Made by OpenAI. Released 2026-04-23. Accessibility not recorded. Cheapest standard rate we hold: $30 per million output tokens, at short context.
What it costs
| Tier | Kind | In | Out | Cached in | From |
|---|---|---|---|---|---|
| batch | text · long | $5 | $23 | $0.50 | seller |
| batch | text · short | $2.50 | $15 | $0.25 | seller |
| fast | text · long | $13 | $75 | $1.25 | seller |
| fast | text · short | $13 | $75 | $1.25 | seller |
| flex | text · long | $5 | $23 | $0.50 | seller |
| flex | text · short | $2.50 | $15 | $0.25 | seller |
| standard | text · long | $10 | $45 | $1 | seller |
| standard | text · short | $5 | $30 | $0.50 | seller |
per million tokens, US dollars. * marks a rate from a single registry or two that disagree — the best available number rather than a confirmed one. Every other row is confirmed against a seller or a licensed index.
What it scores, by house
| Benchmark | House | Score | Evaluated | Scaffold |
|---|---|---|---|---|
| chess puzzles | Epoch AI | 0.3±0.04 | 2026-08-07 | low |
| chess puzzles | Epoch AI | 0.1±0.03 | 2026-08-07 | none |
| ebr bench | Epoch AI | 0.3±0.02 | 2026-07-27 | xhigh |
| frontiermath erdos | Epoch AI | 0.0±0.00 | 2026-08-28 | xhigh |
| frontiermath tier 4 v2 | Epoch AI | 0.7±0.07 | 2026-06-11 | xhigh |
| frontiermath tiers 1 3 v2 | Epoch AI | 0.9±0.02 | 2026-06-11 | xhigh |
| gpqa diamond | Epoch AI | 0.8±0.03 | 2026-08-07 | none |
| gpqa diamond | Epoch AI | 0.9±0.02 | 2026-05-05 | low |
| mirrorcode | Epoch AI | 0.1±0.06 | 2026-08-10 | high |
| mystery game puzzles | Epoch AI | 0.3±0.05 | 2026-08-28 | low |
| mystery game puzzles | Epoch AI | 0.2±0.04 | 2026-08-27 | none |
| mystery game puzzles | Epoch AI | 0.5±0.05 | 2026-07-27 | high |
| mystery game puzzles | Epoch AI | 0.6±0.05 | 2026-07-24 | xhigh |
| otis mock aime 2024 2025near its ceiling | Epoch AI | 0.6±0.07 | 2026-08-07 | none |
| otis mock aime 2024 2025near its ceiling | Epoch AI | 0.8±0.05 | 2026-08-07 | low |
| simpleqa verified | Epoch AI | 0.6±0.02 | 2026-08-27 | xhigh |
| ale bench | collected by Epoch AI | 1128 | — | none |
| ale bench | collected by Epoch AI | 1589 | — | medium |
| ale bench | collected by Epoch AI | 1943 | — | xhigh |
| apex agents | collected by Epoch AI | 0.6 | — | unknown |
| apex agents | collected by Epoch AI | 0.6 | — | xhigh |
| arc aginear its ceiling | collected by Epoch AI | 0.8 | — | low |
| arc aginear its ceiling | collected by Epoch AI | 0.9 | — | high |
| arc aginear its ceiling | collected by Epoch AI | 0.9 | — | medium |
| arc aginear its ceiling | collected by Epoch AI | 0.9 | — | xhigh |
| arc agi 2 | collected by Epoch AI | 0.8 | — | high |
| arc agi 2 | collected by Epoch AI | 0.3 | — | low |
| arc agi 2 | collected by Epoch AI | 0.7 | — | medium |
| arc agi 2 | collected by Epoch AI | 0.8 | — | xhigh |
| blueprint bench 2 | collected by Epoch AI | 0.4 | — | unknown |
| btf3 | collected by Epoch AI | 0.1 | — | Vendor agent SDK |
| btf3 | collected by Epoch AI | 0.1 | — | FutureSearch agent |
| cl bench life | collected by Epoch AI | 0.2 | — | high |
| critpt | collected by Epoch AI | 0.3 | — | high |
| critpt | collected by Epoch AI | 0.0 | — | none |
| critpt | collected by Epoch AI | 0.2 | — | medium |
| critpt | collected by Epoch AI | 0.1 | — | low |
| critpt | collected by Epoch AI | 0.3 | — | xhigh |
| cursorbench | collected by Epoch AI | 0.6 | — | xhigh |
| cursorbench | collected by Epoch AI | 0.5 | — | low |
| cursorbench | collected by Epoch AI | 0.5 | — | medium |
| cursorbench | collected by Epoch AI | 0.6 | — | high |
| deepresearchbench | collected by Epoch AI | 0.5 | — | low |
| deepresearchbench | collected by Epoch AI | 0.5 | — | high |
| deepresearchbench | collected by Epoch AI | 0.5 | — | medium |
| deepswe | collected by Epoch AI | 0.7 | — | mini-swe-agent |
| dtbenchnear its ceiling | collected by Epoch AI | 1.0 | — | xhigh |
| exploitbench | collected by Epoch AI | 0.4 | — | Codex |
| exploitbench | collected by Epoch AI | 0.5 | — | unknown |
| forecastbench | collected by Epoch AI | 61.1 | — | unknown |
| frontiercode | collected by Epoch AI | 0.4 | — | codex |
| gbaeval | collected by Epoch AI | 0.5 | — | unknown |
| gdp pdf | collected by Epoch AI | 0.3 | — | xhigh |
| gso | collected by Epoch AI | 0.4 | — | OpenHands |
| lmca | collected by Epoch AI | 54.3 | — | xhigh |
| osworld 2 | collected by Epoch AI | 0.1 | — | xhigh |
| posttrainbench | collected by Epoch AI | 0.3 | — | Codex CLI |
| proofbench | collected by Epoch AI | 0.5 | — | xhigh |
| rli | collected by Epoch AI | 0.1 | — | unknown |
| scicode | collected by Epoch AI | 0.5 | — | medium |
| scicode | collected by Epoch AI | 0.6 | — | high |
| scicode | collected by Epoch AI | 0.6 | — | xhigh |
| scicode | collected by Epoch AI | 0.5 | — | low |
| scicode | collected by Epoch AI | 0.5 | — | none |
| simplebench | collected by Epoch AI | 0.7 | — | unknown |
| surface evolver bench | collected by Epoch AI | 0.9 | — | high |
| surface evolver bench | collected by Epoch AI | 0.8 | — | medium |
| terminalbench | collected by Epoch AI | 0.8 | — | Capy |
| terminalbench | collected by Epoch AI | 0.7 | — | clnkr |
| terminalbench | collected by Epoch AI | 0.8 | — | Codex CLI |
| terminalbench | collected by Epoch AI | 0.8 | — | Codex |
| terminalbench | collected by Epoch AI | 0.8 | — | NexAU-AHE |
| vending bench 2 | collected by Epoch AI | 7524 | — | unknown |
| webdev arena | collected by Epoch AI | 1458 | — | unknown |
| webdev arena | collected by Epoch AI | 1486 | — | high |
| webdev arena | collected by Epoch AI | 1508 | — | xhigh |
| weirdml | collected by Epoch AI | 0.8 | — | xhigh |
| weirdml | collected by Epoch AI | 0.8 | — | high |
| weirdml | collected by Epoch AI | 0.7 | — | none |
These scores are not comparable between houses and are not added up. A score is dated by when it was evaluated, not when the model was released, and a score without its scaffold is not a measurement — which is why both are printed.
Every rate we hold · Why there is no single best model · Everything from OpenAI