GPT-5.6 Sol vs GPT-6.1 Sol
EBR Bench?
GPT-5.6 Solat max44.8
GPT-6.1 Solat max54.3
Harvey LAB-AA 1.1?
GPT-5.6 Sol—
GPT-6.1 Solat max6.9
SpireClimb?
GPT-5.6 Sol—
GPT-6.1 Solat max24.7
SpireClimb StS1 A20 Heart?
GPT-5.6 Sol—
GPT-6.1 Solat max20
SpireClimb StS2 A10?
GPT-5.6 Sol—
GPT-6.1 Solat max29.3
| Benchmark | GPT-5.6 Sol | GPT-6.1 Sol |
|---|---|---|
| AutomationBench-AA | 51.3 | 62.6 |
| GDPval-AA (work tasks) 2.1 | 46.1 | 47.5 |
| AA Analyst Agent | 47.5 · max | 50 · max |
| Terminal-Bench 4.0 | 14.6 | 48 |
| Terminal-Bench Science 0.1 | 22.4 · max | 58.1 · max |
| AA-Briefcase 1.1 · Elo | 1,241.9 | 1,361.2 |
| AA-LCR (long context) 1.1 | 80.3 | 83.3 |
| DeepSWE 1.1 | 61.1 | 72.3 · max |
| SWE-bench Pro (v2 full) | 95.5 · xhigh | 96.1 · xhigh |
| EBR Bench | 44.8 · max | 54.3 · max |
| Furniture Assembly | 56.7 · max | 80 · max |
| τ³-bench · banking · 1.0.1 | 46.9 · xhigh | — |
| τ³-Banking · AA | 36.5 | — |
| ITBench SRE | 56.2 · max | — |
| Terminal-Bench Hard (version unstated) | 62.9 | — |
| EnterpriseOps-Gym-AA | 42.9 · max | — |
| FrontierCode 1.1 | — | 49.1 · max |
| Harvey LAB-AA 1.1 | — | 6.9 · max |
| SWE-bench Pro (v2 hard) | 82.4 · xhigh | — |
| SpireClimb | — | 24.7 · max |
| SpireClimb StS1 A20 Heart | — | 20 · max |
| SpireClimb StS2 A10 | — | 29.3 · max |
Specs
| Spec | GPT-5.6 Sol | GPT-6.1 Sol |
|---|---|---|
| Input, per 1M | $4 | $2 |
| Output, per 1M | $20 | $10 |
| Cost per task? | $0.50 | $0.21 |
| Speed, tok/s | 85 | 45 |
| Context | 1.05M | 1.05M |
| Weights | Closed | Closed |
| Released | Jul 2026 | Sep 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | Partial |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | Partial |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ✓ |
| Run code | ? | ✓ |
| Use a computer | ? | ✓ |
What breaks when you switch
GPT-5.6 Sol → GPT-6.1 Sol
Remove unsupported parameters›
With reasoning enabled, remove temperature, top_p and top_logprobs, plus Chat Completions logprobs or Responses message.output_text.logprobs includes.
Responses API for tool calling · Chat Completions›
Use Responses for tools; Chat Completions supports this model without tool calling.
No none or minimal reasoning›
The none and minimal reasoning efforts are unsupported; choose low, medium, high, xhigh or max.
More than 272K input tokens›
Prompts above 272K input tokens charge 2× input and cache rates and 1.5× output rates for the entire request.
Cache writes billed at 1.25x›
Cache writes cost 1.25× the uncached input token rate.
Ultrafast prices 6x Standard›
Ultrafast mode costs 6× Standard token prices.
GPT-6.1 Sol → GPT-5.6 Sol
Nothing known.