Codestral 2508 vs Granite 4.0 Micro
Granite 4.0 Micro costs 8.0x less per output token than Codestral 2508 ($0.112 vs $0.90 per million). Updated October 22, 2026; no benchmark has tested both models yet.
- Coding
- No data
- Overall
- No data
- Output $/M
- $0.90
- Coding
- No data
- Overall
- No data
- Output $/M
- $0.112
Codestral 2508 or Granite 4.0 Micro: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Codestral 2508 for a larger context window (256K against 131K tokens).
No benchmark has tested both yet. Codestral 2508 has results on ALE-Bench, Kagi LLM Benchmark and Design Arena: websites; Granite 4.0 Micro on LiveCodeBench (Artificial Analysis run), Terminal-Bench Hard (Artificial Analysis run) and AA Intelligence Index. Compare them on price and context instead.
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
No call
Only Granite 4.0 Micro has a AA Intelligence Index result, 5.1.
Directional only
Web design
Design Arena: websites
No call
Only Codestral 2508 has a Design Arena: websites result, 1012.
Directional only
Long documents
Listed context window
Codestral 2508
Codestral 2508 takes 256K tokens against 131K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Granite 4.0 Micro
Codestral 2508 $0.0008, Granite 4.0 Micro $0.0001 per turn.
List prices
At a glance
- Coding index
- Codestral 2508 not benchmarked, Granite 4.0 Micro not benchmarked.
- Widest category
- None yet: no category has been tested on both models.
- Widest single gap
- None yet: no benchmark has tested both models.
- Output price
- Codestral 2508 $0.90 against Granite 4.0 Micro $0.112 per million output tokens, 8.0x apart.
- Context window
- Codestral 2508 takes 256K tokens, Granite 4.0 Micro 131K.
Benchmark by benchmark
Every board either model has a result on, a category to a fold; bars are each result on a 0 to 100 scale. Only one model is measured on each, so no row names a leader.
CodingDirectional onlySix of these boards feed the Vibe Coding Index3 boards measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | Not measured | 18.0% | Only Granite 4.0 Micro measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 1.5% | Only Granite 4.0 Micro measured |
| ALE-Bench | 137.8 | Not measured | Only Codestral 2508 measured |
| Neither model has a result on the other 36 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 5.1 | Only Granite 4.0 Micro measured |
| Neither model has a result on the other 1 board here | |||
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| τ²-Bench Telecom | Not measured | 12.6% | Only Granite 4.0 Micro measured |
| Neither model has a result on the other 40 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index4 boards measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 33.6% | Only Granite 4.0 Micro measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 5.0% | Only Granite 4.0 Micro measured |
| GPQA Diamond (Epoch AI run) | Not measured | 28.3% | Only Granite 4.0 Micro measured |
| Kagi LLM Benchmark | 32.5% | Not measured | Only Codestral 2508 measured |
| Neither model has a result on the other 18 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| AA Math Index | Not measured | 6.0 | Only Granite 4.0 Micro measured |
| AIME 2025 (Artificial Analysis run) | Not measured | 6.0% | Only Granite 4.0 Micro measured |
| OTIS Mock AIME | Not measured | 2.8% | Only Granite 4.0 Micro measured |
| Neither model has a result on the other 10 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index1 board measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | Not measured | 44.7% | Only Granite 4.0 Micro measured |
| Neither model has a result on the other 17 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| Design Arena: websites | 1012 | Not measured | Only Codestral 2508 measured |
| Design Arena: UI components | 1014 | Not measured | Only Codestral 2508 measured |
| Design Arena: game dev | 967 | Not measured | Only Codestral 2508 measured |
| Design Arena: data viz | 1017 | Not measured | Only Codestral 2508 measured |
| Design Arena: 3D | 1019 | Not measured | Only Codestral 2508 measured |
| Design Arena: all code categories | 1007 | Not measured | Only Codestral 2508 measured |
| Neither model has a result on the other 3 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index2 boards measured
| Benchmark | Codestral 2508 | Granite 4.0 Micro | Leader |
|---|---|---|---|
| IFBench | Not measured | 24.8% | Only Granite 4.0 Micro measured |
| Long context reasoning (AA-LCR) | Not measured | 7.0% | Only Granite 4.0 Micro measured |
| Neither model has a result on the other 44 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
Granite 4.0 Micro costs less
Repository review
50K input + 3K output tokens
Granite 4.0 Micro costs less
Agent loop
200K input + 10K output tokens
Granite 4.0 Micro 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 Codestral 2508 | IBM Granite 4.0 Micro |
|---|---|---|
| Maker | Mistral AI | IBM Granite |
| Weights | No public weights listed | Open weights |
| Released | August 1, 2025 | October 20, 2025 |
| Context window | 256K | 131K |
| Input $/Mtok | $0.30 | $0.017 |
| Output $/Mtok | $0.90 | $0.112 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Codestral 2508 better than Granite 4.0 Micro for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Codestral 2508 for a larger context window (256K against 131K tokens).
Which is cheaper, Codestral 2508 or Granite 4.0 Micro?
Codestral 2508 costs $0.30 per million input tokens and $0.90 per million output tokens; Granite 4.0 Micro costs $0.017 and $0.112. Cost per solved task is known for neither of them.
Which has the bigger context window, Codestral 2508 or Granite 4.0 Micro?
Codestral 2508, with 256K tokens against 131K.
Has any benchmark tested both Codestral 2508 and Granite 4.0 Micro?
No benchmark has tested both yet. Codestral 2508 has results on ALE-Bench, Kagi LLM Benchmark and Design Arena: websites; Granite 4.0 Micro on LiveCodeBench (Artificial Analysis run), Terminal-Bench Hard (Artificial Analysis run) and AA Intelligence Index. Compare them on price and context instead. Data as of October 22, 2026.