GPT-6.2 Sol
92% of Astra's intelligence at a quarter of the price — the value pick of the everyday tier
Benchmark scores
How it is measured →Ranked 5 · bars are scaled to the field maximum for that benchmark.
| Benchmarks | Score | Vs. best in field | Gap |
|---|---|---|---|
| AA Intelligence Index v5 Intelligence index | 57.0 | 88% | −8.0 pts |
| AA Intelligence Index v4.3.2 | — | — | not indexed |
| GPQA Diamond | — | — | not indexed |
| Terminal-Bench 4.0 Terminal-agent tasks | 60.1 | 89% | −7.1 pts |
| LiveBench Design | — | — | not indexed |
| Vals Index | — | — | not indexed |
Assessment
Shipped on 17 November 2026 alongside Astra, replacing GPT-6.1 Sol, whose source row was pulled the same day. AA Intelligence Index v5 reaches 57.26 at max effort, fifth overall and three points above 6.1 Sol's 54.6; Terminal-Bench 4.0 (AA's own run) is 60.1%, four points above 6.1 Sol's 56.1. Pricing holds at $2/$10 with $0.1 cache — a quarter of Astra on both input and output — while time to first answer token drops from 161 seconds on 6.1 to 68, at 85 tok/s. The same-generation Ultrafast tier opens to Pro 200 for the first time. GPQA, Vals and LiveBench read null in AA's record (not run), so they stay blank.
At the same price
The four models closest on blended price, side by side.
| Model | v5 | Blended | tok/s | Score gap |
|---|---|---|---|---|
| Claude Sonnet 5.5Anthropic | 54.5 | $4.00 | 146 | -2.5 |
| Gemini 4 CarbonGoogle | 60.0 | $4.00 | 80 | +3.0 |
| Gemini 4 ArgonGoogle | 50.4 | $4.00 | 65 | -6.6 |
| Kimi K3.1 Pro Open weightsMoonshot AI | 46.0 | $4.00 | 32 | -11.0 |
Compare side by side
Line it up against other models on price, speed and every benchmark.
The comparison tool lines up to 4 models side by side with switchable normalisation — built for purchasing decisions.
Open the comparison tool → Back to leaderboardRelated benchmarks
Data sources & methodology
Full methodology & sources →Every number on this page comes from published evaluations. We run no tests of our own and accept no vendor-submitted scores. Main sources:
Evaluation data in the wider industry also comes from public datasets, leaderboard platforms and their evaluation harnesses, private in-house sets, production logs and user feedback, and synthetic data. This site's current rankings use only the parts that are public, independently checkable and clearly defined.