GPT-6.2 Astra vs Qwen3.6 Max
Full profiles: GPT-6.2 Astra (OpenAI) and Qwen3.6 Max (Alibaba): every benchmark, price and rank.
Updated · Data from 10 sources
As of 19 Nov 2026, GPT-6.2 Astra is ahead of Qwen3.6 Max on quality (68.9 vs 62.0 on the BenchLeader Index), but Qwen3.6 Max is 5.5× cheaper at $2.92 vs $16.00 per 1M tokens blended. Choose GPT-6.2 Astra for coding (76 vs 56), knowledge (74 vs 59), and reasoning (73 vs 61); choose Qwen3.6 Max for agents & tools (72 vs 66).
Key facts (as of 19 Nov 2026)
- GPT-6.2 Astra scores 68.9 on the BenchLeader Index and Qwen3.6 Max scores 62.0 (BenchLeader, 19 Nov 2026).
- GPT-6.2 Astra costs $16.00 per 1M tokens blended (list price, 19 Nov 2026).
- Qwen3.6 Max costs $2.92 per 1M tokens blended (list price, 19 Nov 2026).
- Of the 4 benchmarks both models have, GPT-6.2 Astra leads 4 (BenchLeader, 19 Nov 2026).
- On Arena WebDev, GPT-6.2 Astra is rated 1781 and Qwen3.6 Max 1482 (LMArena, as published by 19 Nov 2026).
Which to pick for…
The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.
| Use case | Pick | Why | Confidence |
|---|---|---|---|
| Coding | GPT-6.2 Astra | 76 vs 56 category score; Arena WebDev 1781 vs 1482 | High |
| Agents & tool use | Qwen3.6 Max | 72 vs 66 category score | Medium |
| Maths | GPT-6.2 Astra | 73 vs 64 category score; OTIS Mock AIME 100.0% vs 91.1% | High |
| Reasoning | GPT-6.2 Astra | 73 vs 61 category score; GPQA Diamond 96.8% vs 87.4% | High |
| Knowledge | GPT-6.2 Astra | 74 vs 59 category score; SimpleQA Verified 78.5% vs 52.0% | High |
| Chat quality (human preference) | – | Only Qwen3.6 Max has an Arena rating (65 category score) | Not comparable |
| Lowest cost | Qwen3.6 Max | $2.92 vs $16.00 per 1M blended; agentic coding session $0.109 vs $0.640 | – |
| High volume on a budget | Neither flagship | A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs | – |
Key differences
- GPT-6.2 Astra (max) leads on quality: 68.9 vs 62.0.
- GPT-6.2 Astra (max) is stronger in coding, knowledge, maths, reasoning.
- Qwen3.6 Max is stronger in agents & tools.
- Qwen3.6 Max is 5.5× cheaper ($2.92 vs $16.00 per 1M blended).
At a glance
| Spec | GPT-6.2 Astra | Qwen3.6 Max |
|---|---|---|
| Developer | OpenAI | Alibaba |
| Released | 17 Nov 2026 | 20 Apr 2026 |
| Context window (tokens) | 1,050,000 | 256,000 |
| Compared at | GPT-6.2 Astra (max) | Qwen3.6 Max |
| Open weights | No | No |
| Price in / out / cached input per 1M | $8.00 / $40.00 / $0.400 | $1.30 / $7.80 / $0.130 |
Category by category
Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.
| Category | GPT-6.2 Astra (max) | Qwen3.6 Max | Leader |
|---|---|---|---|
| Reasoning | 73.1 | 60.8 | GPT-6.2 Astra |
| Coding | 76.1 | 55.6 | GPT-6.2 Astra |
| Agents & tools | 66.4 | 72.1 | Qwen3.6 Max |
| Maths | 72.5 | 64.4 | GPT-6.2 Astra |
| Knowledge | 74.2 | 59.2 | GPT-6.2 Astra |
| Instruction following | – | 74.7 | not comparable |
| Human preference | – | 64.5 | not comparable |
| Multimodal | 71.7 | – | not comparable |
| Long context | – | 65.0 | not comparable |
| Composite | 69.5 | – | not comparable |
On one chart
Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.
Not drawn, measured for fewer than two of these models: Instruction following, Long context, Multimodal, Human preference.
Where sources disagree
Across the 4 benchmarks both have, GPT-6.2 Astra leads 4, Qwen3.6 Max leads 0. Every overall index that rates both puts GPT-6.2 Astra ahead.
| Overall index | GPT-6.2 Astra (max) | Qwen3.6 Max | Leader |
|---|---|---|---|
| BenchLeader Index | 68.9 | 62.0 | GPT-6.2 Astra |
Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).
Price, speed and latency
List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.
| Measure | GPT-6.2 Astra (max) | Qwen3.6 Max |
|---|---|---|
| Blended price per 1M tokens | $16.00 | $2.92 |
| Input / output per 1M | $8.00 / $40.00 | $1.30 / $7.80 |
| Cached input per 1M | $0.400 | $0.130 |
| Price per task: chat reply | $0.015 | $0.0029 |
| Price per task: summarize a 30-page report | $0.120 | $0.020 |
| Price per task: code edit | $0.108 | $0.019 |
| Price per task: agentic coding session | $0.640 | $0.109 |
| Price per task: structured extraction | $0.024 | $0.0042 |
| Agentic coding session, input mostly cached | $0.298 | $0.057 |
| Cost per benchmark task (AA Index) | $4.00 | – |
| Output tokens per task (AA Index) | 41k | – |
| Of which reasoning | 26k | – |
| Time per task (AA Index) | 14 min | – |
| Output speed | 49 tok/s | – |
| Time to first answer | 202.7 s | – |
Each model is compared at its best-ranked reasoning-effort setting. GPT-6.2 Astra (max) scores 68.9 and takes 203 s to its first answer token; GPT-6.2 Astra (xhigh) scores 67.1 in 67 s, GPT-6.2 Astra (medium) scores 65.6 in 4 s. Qwen3.6 Max scores 62.0. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.
Every benchmark
108 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.
| Metric | Publisher | GPT-6.2 Astra (max) | Qwen3.6 Max |
|---|---|---|---|
| BenchLeader Index | BenchLeader | 68.9 | 62.0 |
| Agents & tools score | BenchLeader | 66.4 | 72.1 |
| Coding score | BenchLeader | 76.1 | 55.6 |
| Composite score | BenchLeader | 69.5 | – |
| Knowledge score | BenchLeader | 74.2 | 59.2 |
| Maths score | BenchLeader | 72.5 | 64.4 |
| Multimodal score | BenchLeader | 71.7 | – |
| Reasoning score | BenchLeader | 73.1 | 60.8 |
| Human preference score | BenchLeader | – | 64.5 |
| Instruction following score | BenchLeader | – | 74.7 |
| Long context score | BenchLeader | – | 65.0 |
| Blended price $/M | BenchLeader | $16.00 | $2.92 |
| Output speed | BenchLeader | 49 tok/s | – |
| Time to first answer | BenchLeader | 202.7 s | – |
| Context window | BenchLeader | 1,050,000 tokens | 256,000 tokens |
| Reasoning | |||
| GPQA Diamond | Epoch AI Benchmarking Hub | 96.8% | 87.4% |
| SimpleBench | SimpleBench | – | 63.0% |
| Arena Hard Prompts | Arena (LMArena) | – | 1482 |
| LiveBench 2.0 Reasoning | LiveBench | 91.6% | – |
| GPQA Diamond (AA) | Artificial Analysis | – | 88.8% |
| Humanity's Last Exam (AA) | Artificial Analysis | – | 30.8% |
| HLE-Diamond (AA) | Artificial Analysis | 72.2% | – |
| ARC-AGI-1 | ARC Prize | 98.5% | – |
| ARC-AGI-2 | ARC Prize | 95.8% | – |
| ARC-AGI-3 | ARC Prize | 99.8% | – |
| CritPt | Artificial Analysis | – | 3.7% |
| CritPt v2 hard subset (AA) | Artificial Analysis | 71.5% | – |
| MysteryMechanism | Vals AI | 53.1% | – |
| HLE-Diamond (Scale) | Scale AI SEAL | 66.9%† | – |
| Arena Expert | Arena (LMArena) | – | 1499 |
| Chess Puzzles | Epoch AI Benchmarking Hub | 78.0% | – |
| Mystery Game Puzzles | Epoch AI Benchmarking Hub | 95.0% | 19.0%† |
| LMCA | Epoch AI Benchmarking Hub | – | 42.5% |
| DTBench | Epoch AI Benchmarking Hub | – | 87.2% |
| Coding | |||
| SWE-bench Verified (Epoch) | Epoch AI Benchmarking Hub | – | 76.7% |
| Arena Coding | Arena (LMArena) | – | 1511 |
| Arena WebDev | Arena (LMArena) | 1781 | 1482 |
| LiveBench 2.0 Coding | LiveBench | 91.9% | – |
| SWE-bench (Vals) | Vals AI | – | 72.8% |
| Vibe Code Bench v1.1 | Vals AI | 89.9% | – |
Show all 108 benchmarks
| Metric | Publisher | GPT-6.2 Astra (max) | Qwen3.6 Max |
|---|---|---|---|
| Coding | |||
| Vibe Code Bench 1-100 v1.1 | Vals AI | 55.2% | – |
| SWE Atlas v1.1 | Scale AI SEAL | 60.5%† | – |
| SWE Atlas v1.1: Codebase QnA | Scale AI SEAL | 67.7%† | – |
| SWE Atlas v1.1: Test Writing | Scale AI SEAL | 62.2%† | – |
| SWE Atlas v1.1: Refactoring | Scale AI SEAL | 51.4%† | – |
| MirrorCode | Epoch AI Benchmarking Hub | 67.8% | – |
| Terminal-Bench 4.0 (AA) | Artificial Analysis | 64.1% | – |
| Design Arena Website | Design Arena | 1339† | – |
| FrontierCode 2.0 (AA) | Artificial Analysis | 45.4% | – |
| Agents & tools | |||
| Terminal-Bench | Terminal-Bench | 62.1% | – |
| Arena Agent | Arena (LMArena) | 15.2 | – |
| LiveBench Long Agentic | LiveBench | 70.8% | – |
| LiveBench Browser | LiveBench | 93.0% | – |
| Terminal-Bench Hard | Artificial Analysis | – | 43.9% |
| τ²-Bench Telecom (AA) | Artificial Analysis | – | 95.9% |
| Analyst Agent (AA) | Artificial Analysis | 62.5% | – |
| Finance Agent v2.1 | Vals AI | 60.8% | – |
| Harvey's Legal Agent Benchmark v1.1 | Vals AI | 16.3% | – |
| Legal Research Bench | Vals AI | 42.8% | – |
| Robo-bench | Vals AI | 32.7% | – |
| Tax Agent Bench | Vals AI | 21.8% | – |
| Terminal-Bench 2.0 (Vals) | Vals AI | – | 51.7% |
| Terminal-Bench 4.0 (Vals) | Vals AI | 63.1% | – |
| Terminal-Bench Science | Vals AI | 71.4% | – |
| Arena Agent Task Completion | Arena (LMArena) | 12.4 | – |
| Vending-Bench 2 | Andon Labs | – | 4254.2 |
| SpireClimb | Epoch AI Benchmarking Hub | 41.3% | – |
| AutomationBench | Artificial Analysis | 77.3% | – |
| GDP.pdf (AA) | Artificial Analysis | 29.8% | – |
| Harvey LAB-AA v1.1 | Harvey | 15.0% | – |
| EnterpriseOps-Gym | Artificial Analysis | 29.9% | – |
| AA-Briefcase v1.1 | Artificial Analysis | 1643 | – |
| Design Arena Agentic | Design Arena | 1304 | – |
| Terminal-Bench-Science 0.1 (AA) | Artificial Analysis | 78.1% | – |
| CWE-Bench (AA) | Artificial Analysis | 65.0% | – |
| DeepSec-Bench (AA) | Artificial Analysis | 46.8% | – |
| CyberGym end-to-end (AA) | Artificial Analysis | 0.0% | – |
| MKA (AA) | Artificial Analysis | 52.7% | – |
| Maths | |||
| FrontierMath Tiers 1–3 | Epoch AI Benchmarking Hub | 92.6% | – |
| FrontierMath Tier 4 | Epoch AI Benchmarking Hub | 100.0% | – |
| OTIS Mock AIME | Epoch AI Benchmarking Hub | 100.0% | 91.1% |
| Autoform-Hard | Vals AI | 87.5% | – |
| Arena Maths | Arena (LMArena) | – | 1472 |
| FrontierMath Erdős | Epoch AI Benchmarking Hub | 10.3% | – |
| Knowledge | |||
| SimpleQA Verified | Epoch AI Benchmarking Hub | 78.5% | 52.0% |
| LiveBench 2.0 Data Analysis | LiveBench | 84.1% | – |
| LiveBench Research | LiveBench | 91.5% | – |
| AA-Omniscience v1.1 | Artificial Analysis | 50 | – |
| CorpFin | Vals AI | – | 66.5% |
| PRBench Legal | Scale AI SEAL | 43.1%† | – |
| BioMysteryBench | Vals AI | 80.7% | – |
| CaseLaw v2 | Vals AI | – | 47.9% |
| Excel Modeling Benchmark v1.1 | Vals AI | 76.0% | – |
| DrugDiscoveryBench | Scale AI SEAL | 65.5%† | – |
| EBR-bench | Epoch AI Benchmarking Hub | 92.4% | – |
| AA-Omniscience v1.1: accuracy | Artificial Analysis | 68.6% | – |
| AA-Omniscience v1.1: non-hallucination | Artificial Analysis | 40.9% | – |
| Instruction following | |||
| IFBench | Artificial Analysis | – | 76.6% |
| MultiChallenge | Scale AI SEAL | 73.2%† | – |
| Arena Instruction Following | Arena (LMArena) | – | 1451 |
| Human preference | |||
| Arena Text | Arena (LMArena) | – | 1460 |
| Arena Creative Writing | Arena (LMArena) | – | 1437 |
| Arena Multi-turn | Arena (LMArena) | – | 1467 |
| Arena Longer Queries | Arena (LMArena) | – | 1474 |
| Multimodal | |||
| LiveBench Design | LiveBench | 82.8% | – |
| MMMU-Pro | Artificial Analysis | 89.5% | – |
| Humanity's Sixth Sense | Scale AI SEAL | 63.2%† | – |
| Furniture Assembly | Epoch AI Benchmarking Hub | 89.2% | – |
| VIALS-AA | Handshake AI | 65.8% | – |
| AA-3D | Artificial Analysis | 1318 | – |
| Long context | |||
| AA-LCR v1.1 | Artificial Analysis | – | 80.7% |
| MLCR | Artificial Analysis | 39.4% | – |
| Composite | |||
| Epoch Capabilities Index | Epoch AI Benchmarking Hub | 174.2† | 149.9 |
| LiveBench 2.0 | LiveBench | 84.9% | – |
| AA Intelligence Index v5.0.1 | Artificial Analysis | 62.1 | – |
| AA Intelligence Index v4.3.2 | Artificial Analysis | – | 28.4 |
| AA Progress Index v1 | Artificial Analysis | – | 297.1 |
| Vals Index | Vals AI | 67.7 | – |
| Vals Finance Index | Vals AI | 57.3% | – |
| Vals RSI Index | Vals AI | 29.0% | – |
| Vals Science Index | Vals AI | 68.4% | – |
| Vals Legal Index | Vals AI | 69.7% | – |
| AA Cyber Index | Artificial Analysis | 37.3 | – |
Data as of 2026-11-19. Best configuration of each model; every score links to its source on the model pages.
What changed
- GPT-6.2 Astra (max): BenchLeader Index 67.8 → 68.9.
- GPT-6.2 Astra (max): 6 new results (VIALS-AA, Arena WebDev, Arena Agent…).
- GPT-6.2 Astra (max): BenchLeader Index 67.2 → 67.8.
- GPT-6.2 Astra (max): 2 new results (MKA (AA), Terminal-Bench).
- Qwen3.6 Max: BenchLeader Index 62.8 → 62.3.
- Qwen3.6 Max: BenchLeader Index 63.3 → 62.8.
Related comparisons
- GPT-6.2 Astra vs Claude Fable 5.2
- GPT-6.2 Astra vs Gemini 4 Carbon
- GPT-6.2 Astra vs Claude Opus 5.5
- Qwen3.6 Max vs Claude Fable 5.2
- Qwen3.6 Max vs Gemini 4 Carbon
- Qwen3.6 Max vs Claude Opus 5.5
- GPT-6.1 Terra vs Qwen3.6 Max
- GPT-6.2 Astra vs GPT-6 Astra
Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs
GPT-6.2 Astra vs Qwen3.6 Max: questions
- Is GPT-6.2 Astra better than Qwen3.6 Max?
- GPT-6.2 Astra (max) leads on quality: 68.9 vs 62.0. The BenchLeader Index combines every independent quality benchmark; GPT-6.2 Astra (max) is ahead overall as of 19 Nov 2026, but check the category scores for your use.
- GPT-6.2 Astra vs Qwen3.6 Max: which is cheaper, and will that change?
- Qwen3.6 Max is cheaper today: $2.92 against $16.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
- Is GPT-6.2 Astra or Qwen3.6 Max better for coding?
- GPT-6.2 Astra scores higher in coding (76 vs 56 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1781 for GPT-6.2 Astra against 1482 for Qwen3.6 Max.
- Is GPT-6.2 Astra or Qwen3.6 Max better for agents and tool use?
- Qwen3.6 Max scores higher in agentic tasks (72 vs 66 on the category index, where 50 is average).
- Which is better for long documents?
- GPT-6.2 Astra accepts more context: 1,050,000 against 256,000 tokens for Qwen3.6 Max.
More questions (2)
- Do independent benchmarks agree on GPT-6.2 Astra vs Qwen3.6 Max?
- Across the 4 benchmarks both have, GPT-6.2 Astra leads 4, Qwen3.6 Max leads 0. Every overall index that rates both puts GPT-6.2 Astra ahead.
- When was GPT-6.2 Astra released?
- GPT-6.2 Astra was released on 17 Nov 2026.