Devstral 2 2512 vs o1
Updated October 21, 2026. 3 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick o1 for coding: it scores 61.7 on the Vibe Coding Index against 53.8 for Devstral 2 2512.
Pick Devstral 2 2512 when output tokens cost 30.0x less, or when the context window is larger, 262K against 200K tokens.
- Shared results
- 3
- Devstral 2 2512 only
- 3
- o1 only
- 65
- Like-for-like categories
- 2 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
o1
o1 scores 61.7 against 53.8.
One benchmark behind each score
Intelligence
AA Intelligence Index
o1
o1 leads, 15.2 against 8.6.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Devstral 2 2512 has a AA Agentic Index result, 2.3.
Directional only
Reasoning
GPQA Diamond
No call
Only o1 has a GPQA Diamond result, 73.2%.
Directional only
Math
AIME
No call
Only o1 has a AIME result, 71.5%.
Directional only
Knowledge
MMLU Pro
No call
Only o1 has a MMLU Pro result, 83.5%.
Directional only
Long documents
Listed context window
Devstral 2 2512
Devstral 2 2512 takes 262K tokens against 200K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Devstral 2 2512
Devstral 2 2512 $0.0014, o1 $0.045 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Devstral 2 2512 has the number, $1.19.
Directional only
At a glance
- Coding index
- Devstral 2 2512 53.8 from 1 benchmark, o1 61.7 from 1 benchmark.
- Widest category
- Intelligence, 6.6 points apart on average over 1 shared benchmark.
- Widest single gap
- AA Coding Index: Devstral 2 2512 31.3 against o1 39.7, 8.4 points apart on a 0 to 100 scale.
- Output price
- Devstral 2 2512 $2.00 against o1 $60.00 per million output tokens, 30.0x apart.
- Context window
- Devstral 2 2512 takes 262K tokens, o1 200K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Devstral 2 2512 | Bars | o1 | Ties |
|---|---|---|---|---|
| Coding | 1 | 1 | ||
| Intelligence | 0 | 1 | ||
| All shared boards | 1 | 2 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index11
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| Aider polyglot | Not measured | 61.7% | Only o1 measured |
| SWE-bench Verified | 53.8% | Not measured | Only Devstral 2 2512 measured |
| AA Coding Index | 31.3 | 39.7 | o1 |
| LiveCodeBench (Vals AI) | 51.8% | 50.3% | Devstral 2 2512 |
| SWE-bench Verified (Vals AI) | 62.8% | Not measured | Only Devstral 2 2512 measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 67.9% | Only o1 measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 12.9% | Only o1 measured |
| WeirdML | Not measured | 46.1% | Only o1 measured |
| LMArena Text: coding | Not measured | 1367 | Only o1 measured |
| LMArena Text: software and IT | Not measured | 1366 | Only o1 measured |
| Neither model has a result on the other 28 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| AA Intelligence Index | 8.6 | 15.2 | o1 |
| Epoch Capabilities Index | Not measured | 141.7 | Only o1 measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| AA Agentic Index | 2.3 | Not measured | Only Devstral 2 2512 measured |
| τ²-Bench Telecom | Not measured | 62.6% | Only o1 measured |
| METR time horizons | Not measured | 51.1% | Only o1 measured |
| Neither model has a result on the other 36 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index12 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 73.2% | Only o1 measured |
| MMMU Pro (Vals AI) | Not measured | 77.4% | Only o1 measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 74.7% | Only o1 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 7.0% | Only o1 measured |
| GPQA Diamond (Epoch AI run) | Not measured | 76.8% | Only o1 measured |
| ARC-AGI | Not measured | 30.7% | Only o1 measured |
| SimpleBench | Not measured | 40.1% | Only o1 measured |
| Chess puzzles | Not measured | 10.0% | Only o1 measured |
| VPCT | Not measured | 37.0% | Only o1 measured |
| LMArena Text: hard prompts | Not measured | 1371 | Only o1 measured |
| LMArena Text: expert | Not measured | 1361 | Only o1 measured |
| LMArena Text: hard prompts in English | Not measured | 1372 | Only o1 measured |
| Neither model has a result on the other 9 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| AIME | Not measured | 71.5% | Only o1 measured |
| MATH 500 (Vals AI) | Not measured | 90.4% | Only o1 measured |
| MGSM (Vals AI) | Not measured | 89.3% | Only o1 measured |
| OTIS Mock AIME | Not measured | 73.3% | Only o1 measured |
| FrontierMath, tiers 1 to 3 | Not measured | 14.7% | Only o1 measured |
| MATH level 5 | Not measured | 94.7% | Only o1 measured |
| LMArena Text: math | Not measured | 1388 | Only o1 measured |
| LMArena Text: mathematical work | Not measured | 1390 | Only o1 measured |
| Neither model has a result on the other 5 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index11 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 83.5% | Only o1 measured |
| LegalBench (Vals AI) | Not measured | 80.4% | Only o1 measured |
| TaxEval (Vals AI) | Not measured | 74.3% | Only o1 measured |
| MedQA (Vals AI) | Not measured | 96.5% | Only o1 measured |
| MMLU Pro (Artificial Analysis run) | Not measured | 84.1% | Only o1 measured |
| SimpleQA Verified | Not measured | 41.1% | Only o1 measured |
| GeoBench | Not measured | 80.0% | Only o1 measured |
| LMArena Text: science | Not measured | 1369 | Only o1 measured |
| LMArena Text: medicine and healthcare | Not measured | 1355 | Only o1 measured |
| LMArena Text: legal and government | Not measured | 1368 | Only o1 measured |
| LMArena Text: business and finance | Not measured | 1328 | Only o1 measured |
| Neither model has a result on the other 6 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index25 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| IFBench | Not measured | 70.3% | Only o1 measured |
| Long context reasoning (AA-LCR) | Not measured | 65.0% | Only o1 measured |
| DTBench | Not measured | 74.7% | Only o1 measured |
| LMCA | Not measured | 22.3% | Only o1 measured |
| Fiction.LiveBench | Not measured | 83.3% | Only o1 measured |
| Creative writing (Lech Mazur) | Not measured | 70.2% | Only o1 measured |
| LMArena Text | Not measured | 1366 | Only o1 measured |
| LMArena Text: instruction following | Not measured | 1367 | Only o1 measured |
| LMArena Text: creative writing | Not measured | 1347 | Only o1 measured |
| LMArena Vision | Not measured | 1168 | Only o1 measured |
| LMArena Text: English | Not measured | 1372 | Only o1 measured |
| LMArena Text: non-English | Not measured | 1358 | Only o1 measured |
| LMArena Text: Chinese | Not measured | 1394 | Only o1 measured |
| LMArena Text: German | Not measured | 1337 | Only o1 measured |
| LMArena Text: Japanese | Not measured | 1346 | Only o1 measured |
| LMArena Text: Korean | Not measured | 1396 | Only o1 measured |
| LMArena Text: French | Not measured | 1344 | Only o1 measured |
| LMArena Text: Spanish | Not measured | 1345 | Only o1 measured |
| LMArena Text: Russian | Not measured | 1356 | Only o1 measured |
| LMArena Text: multi-turn | Not measured | 1355 | Only o1 measured |
| LMArena Text: longer queries | Not measured | 1378 | Only o1 measured |
| LMArena Text: writing and language | Not measured | 1360 | Only o1 measured |
| LMArena Text: entertainment and media | Not measured | 1358 | Only o1 measured |
| LMArena Vision: English | Not measured | 1162 | Only o1 measured |
| LMArena Vision: Chinese | Not measured | 1235 | Only o1 measured |
| Neither model has a result on the other 21 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Devstral 2 2512 | o1 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Devstral 2 2512 costs less
Repository review
50K input + 3K output tokens
Devstral 2 2512 costs less
Agent loop
200K input + 10K output tokens
Devstral 2 2512 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Mistral Devstral 2 2512 | OpenAI o1 |
|---|---|---|
| Maker | Mistral AI | OpenAI |
| Weights | Open weights | No public weights listed |
| Released | December 9, 2025 | December 17, 2024 |
| Context window | 262K | 200K |
| Input $/Mtok | $0.40 | $15.00 |
| Output $/Mtok | $2.00 | $60.00 |
| Coding index | 53.8 | 61.7 |
| Independent measurements behind it | 1 | 1 |
| Cost per solved task | $1.19 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Devstral 2 2512 better than o1 for coding?
Pick o1 for coding: it scores 61.7 on the Vibe Coding Index against 53.8 for Devstral 2 2512. Pick Devstral 2 2512 when output tokens cost 30.0x less, or when the context window is larger, 262K against 200K tokens.
Which is cheaper, Devstral 2 2512 or o1?
Devstral 2 2512 costs $0.40 per million input tokens and $2.00 per million output tokens; o1 costs $15.00 and $60.00. Cost per solved task is known for Devstral 2 2512 only.
Which has the bigger context window, Devstral 2 2512 or o1?
Devstral 2 2512, with 262K tokens against 200K.
How many benchmarks is the Devstral 2 2512 vs o1 comparison based on?
3 benchmarks have a result for both models as of October 21, 2026. Devstral 2 2512's coding index rests on 1 independent benchmark and o1's on 1. Where only one model has a result, the page says so instead of naming a winner.