Across 13 comparable benchmarks, GPT-6.2 Astra has better reported results on 12 and worse on 1.
Moderate evidence13 of 13 shared results countedHow we compare
Price basis: Standard rates vs listed rates. The price section below shows the applicable periods.
| Fact | Mistral Large 4 | GPT-6.2 Astra |
|---|---|---|
| Context window | 1M tokens | 1.05M tokens |
| Input / 1M | $0.700 | $8.00 |
| Output / 1M | $2.10 | $40.00 |
| Blended (4:1) | $0.526 | $8.93 |
| Status | Preview | Available |
| Released | October 6, 2026 | November 17, 2026 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 10 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
GPT-6.2 Astra ahead on 3 of 3 counted
GPT-6.2 Astra ahead on 2 of 2 counted
GPT-6.2 Astra ahead on 2 of 2 counted
GPT-6.2 Astra ahead on 1 of 1 counted
GPT-6.2 Astra ahead on 3 of 4 counted
No shared comparable results for Abstract reasoning, Computer & tool use, Research & search, Long context, Cybersecurity.
Published scores, with matching modes where available. Gaps show Mistral Large 4 relative to GPT-6.2 Astra. Reading the results
| Benchmark | Mistral Large 4 | GPT-6.2 Astra | Gap |
|---|---|---|---|
| FrontierCode 2.0Repository Engineering · Weighted rubric score | 8.8Thinking High · tools + web | 49.0≠Extra-High · tools | 0.18×#30 vs #2 of 36 |
| Terminal-Bench 4.0Agentic Development · Resolution rate (%) | 28.3Thinking · tools | 64.1≠Extra-High · tools | −35.8 pts#40 vs #5 of 90 |
| Vibe Code Bench v1.1Code Generation & Editing · Pass rate (%) | 78.4Thinking High · tools | 89.9≠Max · tools | −11.5 pts#36 vs #9 of 121 |
| FrontierMath Tier 4 v2Mathematics · Accuracy (verification_code) | 26.8Thinking High · tools | 100.0≠Max · tools | −73.2 pts#40 vs #1 of 73 |
| FrontierMath v2Mathematics · Accuracy (verification_code) | 58.6Thinking High · tools | 92.6≠Max · tools | −34.0 pts#46 vs #4 of 85 |
| MysteryMechanismScientific Reasoning · Accuracy (%) | 12.2Thinking High · tools | 53.2≠Max · tools | −41.0 pts#32 vs #1 of 32 |
| BioMysteryBenchBiology & Genomics · Accuracy (%) | 67.0Thinking High · tools | 80.7≠Max · tools | −13.7 pts#21 vs #5 of 28 |
| MMMU-ProVisual Understanding · Accuracy | 76.4Thinking · no tools | 89.5≠Max · no tools | −13.1 pts#55 vs #3 of 136 |
| EMB (Excel Modeling)Finance · Accuracy (%) | 56.2Thinking High · tools | 76.0≠Max · tools | −19.8 pts#49 vs #3 of 78 |
| Harvey's Legal Agent BenchmarkLegal · Accuracy (%) | 15.8Thinking High · tools | 16.3≠Max · tools | −0.4 pts#6 vs #5 of 52 |
| Legal Research BenchLegal · Score | 31.7Thinking High · tools | 42.8≠Max · tools + web | 0.74×#52 vs #29 of 91 |
| Tax Agent BenchFinance · Accuracy (%, rubric × citation quality) | 63.3Thinking High · tools | 21.8≠Max · tools + web | +41.4 pts#14 vs #52 of 74 |
| AA Intelligence Index v5.0.1Capability Indices · Index (0-100) | 25.3Thinking · tools | 62.1≠Max · task tools | 0.41×#39 vs #2 of 62 |
Repository Engineering · Weighted rubric score
Agentic Development · Resolution rate (%)
Code Generation & Editing · Pass rate (%)
Mathematics · Accuracy (verification_code)
Mathematics · Accuracy (verification_code)
Scientific Reasoning · Accuracy (%)
Biology & Genomics · Accuracy (%)
Visual Understanding · Accuracy
Finance · Accuracy (%)
Legal · Accuracy (%)
Legal · Score
Finance · Accuracy (%, rubric × citation quality)
Capability Indices · Index (0-100)
A dash means no result is available.
| Benchmark | Mistral Large 4 | GPT-6.2 Astra |
|---|---|---|
| FrontierMath Erdős | — | 10.3 |
| GPQA Diamond | — | 97.0 |
| HLE-Diamond | — | 72.2 |
| ProofBench v1.1 (Lean 4) | 10.0 | — |
| Terminal-Bench-Science 0.1 | — | 78.6 |
| AA Cyber Index | 49.5 | — |
| Code Migration | 30.6 | — |
| DeepSWE | 61.7 | — |
| FrontierCode 1.1 | 39.4 | — |
| IOI (Vals v2) | 45.3 | — |
| SWE-Atlas-QnA | 59.4 | — |
| Terminal-Bench 4.1 | 27.6 | — |
| Vibe Code Bench 1-100 | 14.3 | — |
| WeirdML v3 | 0.044 | — |
| AutomationBench-AA | — | 77.2 |
| Agent Arena Index | — | 1291 |
| ECI | — | 174 |
| LiveBench | 71.8 | — |
| LiveBench 2.0 | 47.7 | — |
| Vals Index v2.1 | 48.0 | — |
| Vals Index v2.2 | — | 67.7 |
| Vals RSI Index | — | 29 |
| EBR-bench | — | 92.4 |
| GDP.pdf | — | 29.8 |
| AA-AnalystAgent | — | 62.5 |
| BenchCAD 2.1 Core | — | 82.5 |
| Finance Agent v2 | 54.7 | — |
| Harvey LAB-AA v1.1 (All-pass) | — | 15.0 |
| ARC-AGI-1 | — | 98.5 |
| ARC-AGI-2 | — | 95.8 |
| ARC-AGI-3 | — | 100.0 |
| SimpleBench | — | 77.3 |
| SimpleQA Verified | — | 78.5 |
| HealthBench Professional | — | 66.3 |
| MedCode | 40.7 | — |
| MedScribe | 80.4 | — |
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
Mistral Large 4 costs94%lessthan GPT-6.2 Astra per million tokens — 17× cheaper.Using standard rates vs listed rates at this token mix.
Mistral Large 4: Standard pricing. Mistral official pricing (docs.mistral.ai/inference/pricing) verified 2026-11-13; price reduced from $1.36 Mistral official pricing (docs.mistral.ai/inference/pricing) verified 2026-11-13; price reduced from $4.18
GPT-6.2 Astra: Standard pricing. context_length <= 272000 Official OpenAI per-million-token pricing. Full request uses the applicable input-token band. context_length > 272000
| Price / 1M tokens | Mistral Large 4 | GPT-6.2 Astra | Ratio |
|---|---|---|---|
| Input | $0.70 | $8.00 | 0.09× |
| Output | $2.10 | $40.00 | 0.05× |
| Cached input | $0.07 | $0.40 | 0.18× |
| Cache write | — | $10.00 | — |
| Blended | $0.526 | $8.93 | 0.06× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
Mistral Large 4 relative to GPT-6.2 Astra
Summary: 2 of 10 shared benchmarks have matching recorded settings and sources. The average is a geometric mean of 2 positive-cost ratios, one equal weight per benchmark. This benchmark set can include related tests and indexes; it is not a workload cost estimate.
Mistral Large 4 2.6× cheaper
Mistral Large 4 1.1× cheaper
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Mistral Large 4 2.6× cheaper
Mistral Large 4 1.1× cheaper
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Settings or sources differ
Effort, protocol, aggregation and source match; other run conditions may differ.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Effort, protocol, aggregation and source match; other run conditions may differ.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
Settings or sources differ or are not recorded; interpret the comparison with care.
Mistral Large 4
GPT-6.2 Astra
| Benchmark | Mistral Large 4 | GPT-6.2 Astra |
|---|---|---|
| Agent Arena Index | Not reported | $3.2842 / task Median · Agent Arena Index rating (index points): 1291 Max (With Tools + Internet) |
| ARC-AGI-1 | Not reported | 2 more settings |
| ARC-AGI-2 | Not reported | 2 more settings |
| BenchCAD 2.1 Core | Not reported | $2.89 / task Average · Core mean geometry/circuit score (0–100): 74.6 Thinking High (With Tools) $0.98 / task Average · Core mean geometry/circuit score (0–100): 42.6 Thinking Low (With Tools) 2 more settings$1.94 / task Average · Core mean geometry/circuit score (0–100): 68.5 Thinking Medium (With Tools) $3.36 / task Average · Core mean geometry/circuit score (0–100): 77.6 Extra-High (With Tools) |
| Code Migration | Not reported | |
| Finance Agent v2 | Not reported | |
| IOI (Vals v2) | Not reported | |
| LiveBench 2.0 | $0.6726 / task Average · Overall score (equal-weight mean of 7 suite scores, 0-100): 47.7 Thinking High (With Tools) | Not reported |
| MedCode | Not reported | |
| MedScribe | Not reported | |
| ProofBench v1.1 (Lean 4) | Not reported | |
| Terminal-Bench 4.1 | Not reported | |
| Terminal-Bench-Science 0.1 | Not reported | |
| Vals Index v2.1 | Not reported | |
| Vals Index v2.2 | Not reported | |
| Vibe Code Bench 1-100 | Not reported | |
| WeirdML v3 | $28.7748 / task Average · Official score (0-1; higher is better): 0.044 Thinking High (With Tools) | Not reported |
| Attribute | Mistral Large 4 | GPT-6.2 Astra |
|---|---|---|
| Availability | Preview | Available |
| Provider | MistralAI | OpenAI |
| Model type | Multimodal model | Reasoning model |
| First release | October 6, 2026 | November 17, 2026 |
| Context window | 1M tokens | 1.05M tokens |
| Max output | — | 128K tokens |
| Total parameters | 1.1T | — |
| Active parameters | 49B | — |
| Architecture | Mixture of Experts (MoE) | Undisclosed |
| Documented capabilities | Reasoning, Multilingual, Function calling, Tool use, Structured output, Long context, Coding | Reasoning, Coding, Function calling, Tool use, Structured output, Computer use, Web search, Long context |
| Knowledge cutoff | — | 2026-05-18 |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
| Weights | Closed / API only | Closed / API only |
| Commercial use | Proprietary | Proprietary |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Mistral Large 4 or GPT-6.2 Astra? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.