Codestral 2508
mistral/codestral-2508От Mistral · семейство: codestral · выпуск 2025-08-01 · дата знаний: 2025-05
Prices in USD per 1M tokens. Unknown means the provider does not publish per-token pricing.
Возможности
Model fit scores
0–100 · higher is betterThese scores reward declared capabilities, context size, price and provider availability — they are not benchmark results. Use them as a directional signal alongside your own evaluation.
Coding73
- Tool calling40/40
- Structured output20/20
- Reasoning0/10
- Context window (100K → 1M)8/20
- Provider availability5/10
Agents80
- Tool calling35/35
- Structured output25/25
- Reasoning0/15
- Output token limit15/15
- Provider availability5/10
JSON / structured output98
- Structured output / JSON mode50/50
- Tool calling20/20
- Temperature control10/10
- Price-friendly for high-volume18/20
Cost efficiency60
- Headline price (log-scaled)60/95
- Has prompt-cache pricing0/5
Long context60
- Context window (100K → 2M)50/90
- Has published price for full window10/10
Production-readiness85
- Number of independent providers25/40
- Has published per-token price20/20
- Context window ≥ 8K15/15
- No data inconsistencies across providers10/10
- Official model (not derivative)15/15
Cost Efficiency Index
Open full calculator →Estimated cost using the recommended provider's headline rate. Each scenario fixes average input/output tokens — the assumptions are shown in the third column.
| Scenario | Cost | Assumption |
|---|---|---|
RAG answer per 1,000 RAG answers | $1.95 < $0.01 per request | 5K input tokens (query + 4 retrieved chunks of ~1K each) and a 500-token answer. Typical SaaS knowledge-base bot. |
Support ticket triage per 10,000 tickets | $3.90 < $0.01 per request | 1K input tokens (ticket body + system prompt) and a 100-token JSON classification reply. High-volume customer support. |
Data extraction per 1,000 documents | $1.05 < $0.01 per request | 2K input tokens (a single document page) and a 500-token JSON extraction. ETL / invoice / form pipelines. |
Code review per 1,000 PRs | $3.30 < $0.01 per request | 8K input tokens (diff + surrounding files) and a 1K-token review comment. PR-bot workloads. |
Agent step per 1,000 steps | $4.14 < $0.01 per request | 12K input tokens (long-running tool history) and a 600-token tool-call decision. Cost per agent step. |
Детализация цен
Рекомендованная цена от openrouter · mistralai/codestral-2508
Доступна у 5 провайдеров
| Провайдер | ID модели провайдера | Вход / 1M | Выход / 1M | Контекст | Выпуск |
|---|---|---|---|---|---|
| OpenRouter openrouter | mistralai/codestral-2508 | $0.300 | $0.900 | 256K | 2025-08-01 |
| NanoGPT nano-gpt | mistralai/codestral-2508 | $0.300 | $0.900 | 256K | 2025-08-01 |
| Kilo Gateway kilo | mistralai/codestral-2508 | $0.300 | $0.900 | 256K | 2025-08-01 |
| Cortecs cortecs | codestral-2508 | $0.300 | $0.900 | 256K | 2025-07-30 |
| LLM Gateway llmgateway | codestral-2508 | $0.300 | $0.900 | 256K | 2025-07-30 |
Frequently asked questions
How much does Codestral 2508 cost?
Codestral 2508 costs $0.300 per 1M input tokens and $0.900 per 1M output tokens, sourced from openrouter. Cache reads, audio tokens and >200K-context tiers (where applicable) are listed in the Pricing detail block above.
What is the context window of Codestral 2508?
Codestral 2508 has a context window of 256K tokens, with a max output of 256K tokens per reply. This is the total combined size of prompt + completion.
Does Codestral 2508 support tool calling?
Yes. Codestral 2508 supports tool calling (function calling). This makes it suitable for production agent and automation workloads where the model has to invoke external functions reliably.
Does Codestral 2508 support structured output / JSON mode?
Yes. Codestral 2508 supports structured output / JSON-schema-constrained decoding. This makes it suitable for production agent and automation workloads where the model has to invoke external functions reliably.
Can Codestral 2508 accept image input?
No. Codestral 2508 only accepts text as input. If you need image input, see our /capabilities/vision list for current vision-capable models.
Is Codestral 2508 open-weight?
Yes. Codestral 2508's weights are publicly available, so you can self-host or fine-tune. Note that open weights ≠ open source — the training data and code are typically not released.
What are the best alternatives to Codestral 2508?
If Codestral 2508 doesn't fit, consider Mistral Nemo Instruct 2407, Mistral Nemo, Mistral Large 2.1. Each one targets the same use case — see the Related links below for direct head-to-head pages.
Where does this data come from?
All numbers come from the public models.dev API and are normalised into a single canonical model record. We re-pull daily and write any changes (price, context, capability) to the /changelog page.
Explore more
More Mistral models
- Mistral Nemo Instruct 2407$0.02 in / $0.04 out
- Mistral Nemo$0.15 in / $0.15 out
- Mistral Large 2.1$2.00 in / $6.00 out
- Mistral Large 3$0.50 in / $1.50 out
- Mistral Small 3.2 24B Instruct (2506)$0.10 in / $0.31 out
Capability lists this model is in
Последнее обновление:
Data is sourced from models.dev and normalized for comparison. Prices and capabilities may change. Always verify critical production decisions with the provider's official documentation.