OpenAI称两项API设置将GPT-5.6 Sol的ARC-AGI-3成绩提升至38.3%

#### Foundation #### Leaderboards #### Benchmark #### Prize #### Research #### Content ARC Prize Verified # GPT-5.6 Sol GPT-5.6 Sol is the standout model of the GPT-5.6 family. Sol at max reasoning effort is the only performant model (as of July 2026) averaging 13.33% on Public and 7.78% on Semi-Private. It is the first model to win an ARC-AGI-3 public game (ft09, 87%). Sol is able to read an unfamiliar scene correctly and in the game's own vocabulary. It treats a failed hypothesis as a reason to re-plan rather than thrash. Most agent failures are upstream of the code they write or the action they take. Sol is able to perform on ARC-AGI not because it executes better, but because it correctly orients itself in a new environment first. Related models: GPT-5.6 Terra, GPT-5.6 Luna [...] | 5783df64 | ✓ | ✓ | ✓ | ✓ | ✓ | | 5833af48 | ✓ | ✓ | ✓ | ✓ | ✓ | | 58743b76 | ✓ | ✓ | ✓ | ✓ | ✓ | | 58e15b12 | ✓ | ✓ | ✓ | ✓ | ✓ | | 59341089 | ✓ | ✓ | ✓ | ✓ | ✓ | | 5a5a2103 | ✓ | ✓ | ✓ | ✓ | ✓

## Efficient by default, maximum performance on demand GPT‑5.6 Sol is our best coding model yet. On the Artificial Analysis Coding Agent Index, GPT‑5.6 Sol with max reasoning sets a new state of the art at 80, 2.8 points above Fable 5, while using less than half the output tokens, taking less than half the time, and costing about one-third less. That advantage extends across the family: Terra performs just above Fable 5, while Luna outperforms Opus 4.8; each does so in roughly one-third of the time, with about half as many output tokens, and at approximately one-quarter the estimated cost. It also sets new state-of-the-art results on Terminal‑Bench 2.1 and DeepSWE, which test complex command-line workflows and long-horizon engineering in real codebases.

GPT-5.6 Sol on the ARC-AGI-3 Public Set _With the official harness, GPT‑5.6 Sol scored 13.3% on the ARC-AGI-3 public set. With retained reasoning and compaction, it scored 38.3%. Scores measure Relative Human Action Efficiency (__RHAE_⁠(opens in a new window)_)_—_a metric comparing model performance to a human baseline. Based on__official gameplay logs_⁠(opens in a new window)_, we estimate the average human tester scored 48%. Models are not told how they will be scored, and cannot see their score throughout_—_actions only return a text representation of each frame and what level they are on._ ## ARC-AGI-3 [...] GPT‑5.6 Sol has solved longstanding open problems in mathematics like the cycle double cover conjecture⁠(opens in a new window) and beaten games like Pokémon FireRed. But on ARC-AGI-3, a benchmark of 2D puzzle games, GPT‑5.6 Sol scored just 7.8%, and GPT‑5.5 could barely play the games at all, scoring a paltry 0.4%. Were 2D puzzle games unusually difficult for our models? Or

Drew Bredvick / @dbredvick: Your harness matters a lot actually @teortaxestex: Nobody is ahead of OpenAI on post-training Tibo / @thsottiaux: Terrific work by @ilanbigio and @sandersted on the investigation and post. Seems like it was a lot of fun. Take care of your harnesses out there, your model will thank you. @scaling01: bruh Peter Steinberger / @steipete: lol did nobody at Anthropic stop for a second and wonder why the numbers looked this absurd before posting the “victory”-tweet? [...] | | | OpenAI: | … More: The Decoder and RuntimeWire X: @thsottiaux, @openai, @\_xjdr, @realchillben, @sama, @openai, @kimmonismus, @emollick, @daniel\_mac8, @dbredvick, @teortaxestex, @thsottiaux, @scaling01, @steipete, @ns123abc, @charliermarsh, @rohanpaul\_ai, and @chrisgpt. Mastodon: @[email protected] More: Matthias Bastian / The Decoder: OpenAI claims GPT-5.6 Sol beats Opus 5 on ARC-AGI-3 with its latest API and two additional settings Ryan Merket / RuntimeWire: O

Log inSign up ## Post user avatar GPT-5.6 Sol has been used to solve open problems in mathematics. So why was it struggling with ARC-AGI-3, a benchmark of 2D puzzle games? We investigated. The harness was not letting it remember what it had learned. We found that enabling two API settings tripled our scores with 6x fewer output tokens. 00:00 11:57 PM · Jul 29, 20261.1MViews user avatar OpenAI @OpenAI Jul 29 ARC-AGI-3 tests how well models can learn unfamiliar 2D games without instructions. The standard harness discarded GPT-5.6 Sol’s reasoning after each move and dropped earlier actions as the context filled up. The model had to keep starting over. 00:00 100K user avatar OpenAI @OpenAI Jul 29 [...] 100K user avatar OpenAI @OpenAI Jul 29 We implemented the harness with the Responses API and turned on: → Retained reasoning → Context compaction On the public set, GPT-5.6 Sol’s score rose 188% while using 6x fewer output tok

Log inSign up ## Post user avatar Tibo @thsottiaux Terrific work by @ilanbigio and @sandersted on the investigation and post. Seems like it was a lot of fun. Take care of your harnesses out there, your model will thank you. user avatar Tibo @thsottiaux Jul 29 Turns out GPT-5.6 Sol is actually SoTA on ARC-AGI-3. Just took two setting changes. You just have to allow it to reason and work over multiple context windows with the help of our canonical compaction implementation. openai.com/index/how-two-… 1:22 AM · Jul 30, 2026516.3KViews user avatar ffin @ffinbuilds Jul 30 Please remove 5h limits or make it so we can set our own instead! The workflow disruption is very painful. 1.1K user avatar X Girls @thesoragirls Jul 30 [...] 1.1K user avatar X Girls @thesoragirls Jul 30 Tibo preparing a PSA for people to take care of their harnesses 😂 GIF 2.8K user avatar xy @xyc3yf Jul 30 Please remove 5h li