Skip to content
llmwaves

GPT-5.6 Sol

OpenAI · released Jul 9, 2026

ProprietaryReasoningTool useStructured outputVision

GPT-5.6 Sol is the flagship model in OpenAI's GPT-5.6 series. It is suited for complex reasoning, coding, and agentic workflows, and is particularly strong at command-line and multi-step coding tasks...

Specification

Context window
1.05M
Max output
128K
Knowledge cutoff
Feb 16, 2026
Parameters
Undisclosed
Licence
Proprietary
Serving providers
3
Moderated
Yes
Uptime
100.0%

Intelligence

60.9

99th percentile

Coding

77.4

Coding Index

Agentic

57.8

Agentic Index

Output speed

46 t/s

Median across providers

Latency

2.38s

Time to first token

Cost per task

$2.65

Estimated

Benchmarks

Where the score comes from

The Intelligence Index is a composite. These are the underlying evaluations this model was actually measured on.

Evaluation scores

Percentage correct · higher is better

  • GPQA Diamond
    94.1%
  • τ²-bench (Telecom)
    85.1%
  • AA-LCR (long context)
    77.7%
  • IFBench
    72.7%
  • Terminal-Bench Hard
    65.9%
  • SciCode
    56.1%
  • Humanity's Last Exam
    49.5%

An evaluation missing from this list was not run for this model — it is not a zero.

View as table
EvaluationScore
GPQA Diamond94.1%
τ²-bench (Telecom)85.1%
AA-LCR (long context)77.7%
IFBench72.7%
Terminal-Bench Hard65.9%
SciCode56.1%
Humanity's Last Exam49.5%

Against its peers

Intelligence Index · this model highlighted, nearest peers in grey

Peers are the models sitting closest on the Intelligence Index — the set you would realistically choose between.

View as table
ModelIntelligence
Claude Opus 563.1
Claude Fable 562.1
GPT-5.6 Sol60.9
Kimi K359.7
Qwen3.8 Max58.1
Claude Opus 4.857.3
Muse Spark 1.256.8
GPT-5.6 Terra56.6

Percentile among all indexed models

Intelligence99th
Coding99th
Agentic96th
Terminal work100th
Arena Elo99th

Pricing

What it costs to run

List prices per million tokens, plus what one representative task works out to.

List price

Input / 1M tokens
$5
Output / 1M tokens
$30
Cached input / 1M
$0.5
Blended 3:1
$11.25

One task, estimated

$2.65

Input tokens
50,000
Output tokens
80,000
Profile
Reasoning

Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).

Serving providers

Speed and latency figures are medians across these providers, so a widely-served model reports a blend rather than any single endpoint.

Arena

Head-to-head generation quality

Elo from pairwise judgements, broken out by the kind of thing the model was asked to build.

Overall Elo
1,380
Win rate
62.7%
Strongest at
Game development
Tournaments
1,265

Elo by category

Dot position on a 1,325–1,400 scale · Elo has no meaningful zero

Agent categories (full-stack apps, mobile apps) are only scored for models tested in agent harnesses.

View as table
CategoryEloWin rate
Game development138062.7%
UI components136559.6%
SVG136564.7%
3D scenes136459.3%
Websites134459.8%
Data visualisation134357.9%