Devstral 2 2512 vs Kimi K2 0905
Updated October 21, 2026. 1 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Neither model has a coding score yet, so the two cannot be ranked on coding.
Devstral 2 2512 has the edge where output tokens cost 1.3x less, and where the benchmark that matters is AA Intelligence Index, where it leads.
- Shared results
- 1
- Devstral 2 2512 only
- 5
- Kimi K2 0905 only
- 15
- Like-for-like categories
- 1 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, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Kimi K2 0905
Kimi K2 0905 leads, 15.3 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
Web design
Design Arena: websites
No call
Only Kimi K2 0905 has a Design Arena: websites result, 1089.
Directional only
Long documents
Listed context window
No call
Both take 262K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Devstral 2 2512
Devstral 2 2512 $0.0014, Kimi K2 0905 $0.0019 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 not benchmarked, Kimi K2 0905 not benchmarked.
- Widest category
- Intelligence, 6.7 points apart on average over 1 shared benchmark.
- Widest single gap
- AA Intelligence Index: Devstral 2 2512 8.6 against Kimi K2 0905 15.3, 6.7 points apart on a 0 to 100 scale.
- Output price
- Devstral 2 2512 $2.00 against Kimi K2 0905 $2.50 per million output tokens, 1.3x apart.
- Context window
- Both take 262K tokens.
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 | Kimi K2 0905 | Ties |
|---|---|---|---|---|
| Intelligence | 0 | 1 | ||
| All shared boards | 0 | 1 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index8 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| SWE-bench Verified | 53.8% | Not measured | Only Devstral 2 2512 measured |
| AA Coding Index | 31.3 | Not measured | Only Devstral 2 2512 measured |
| LiveCodeBench (Vals AI) | 51.8% | Not measured | Only Devstral 2 2512 measured |
| SWE-bench Verified (Vals AI) | 62.8% | Not measured | Only Devstral 2 2512 measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 61.0% | Only Kimi K2 0905 measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 23.5% | Only Kimi K2 0905 measured |
| WeirdML | Not measured | 36.7% | Only Kimi K2 0905 measured |
| ALE-Bench | Not measured | 267.1 | Only Kimi K2 0905 measured |
| Neither model has a result on the other 30 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| AA Intelligence Index | 8.6 | 15.3 | Kimi K2 0905 |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| AA Agentic Index | 2.3 | Not measured | Only Devstral 2 2512 measured |
| τ²-Bench Telecom | Not measured | 73.4% | Only Kimi K2 0905 measured |
| Neither model has a result on the other 37 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 76.7% | Only Kimi K2 0905 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 6.4% | Only Kimi K2 0905 measured |
| Neither model has a result on the other 19 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| AA Math Index | Not measured | 57.3 | Only Kimi K2 0905 measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 57.3% | Only Kimi K2 0905 measured |
| Neither model has a result on the other 11 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | Not measured | 81.9% | Only Kimi K2 0905 measured |
| Neither model has a result on the other 16 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1089 | Only Kimi K2 0905 measured |
| Design Arena: all code categories | Not measured | 1098 | Only Kimi K2 0905 measured |
| Neither model has a result on the other 7 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | Leader |
|---|---|---|---|
| IFBench | Not measured | 41.7% | Only Kimi K2 0905 measured |
| Long context reasoning (AA-LCR) | Not measured | 53.0% | Only Kimi K2 0905 measured |
| Fiction.LiveBench | Not measured | 66.7% | Only Kimi K2 0905 measured |
| Neither model has a result on the other 43 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Devstral 2 2512 | Kimi K2 0905 | 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 | MoonshotAI Kimi K2 0905 |
|---|---|---|
| Maker | Mistral AI | Moonshot AI |
| Weights | Open weights | Open weights |
| Released | December 9, 2025 | September 4, 2025 |
| Context window | 262K | 262K |
| Input $/Mtok | $0.40 | $0.60 |
| Output $/Mtok | $2.00 | $2.50 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 1 | 0 |
| Cost per solved task | $1.19 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Devstral 2 2512 better than Kimi K2 0905 for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Devstral 2 2512 has the edge where output tokens cost 1.3x less, and where the benchmark that matters is AA Intelligence Index, where it leads.
Which is cheaper, Devstral 2 2512 or Kimi K2 0905?
Devstral 2 2512 costs $0.40 per million input tokens and $2.00 per million output tokens; Kimi K2 0905 costs $0.60 and $2.50. Cost per solved task is known for Devstral 2 2512 only.
Which has the bigger context window, Devstral 2 2512 or Kimi K2 0905?
Neither: both take 262K tokens.
How many benchmarks is the Devstral 2 2512 vs Kimi K2 0905 comparison based on?
1 benchmarks have a result for both models as of October 21, 2026. Devstral 2 2512's coding index rests on 1 independent benchmark and Kimi K2 0905's on 0. Where only one model has a result, the page says so instead of naming a winner.