impulse

Mistral Medium 3.5

Open weights

Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI.

API id mistral-medium-3-5Released 30 Apr 2026 (Apr 2026)
A newer model in this line is out: Mistral Medium 2604.
Mistral AI (some versions only) retires it on 30 Apr 2027 (in 163 days). 2 hosts keep serving it.
Input / 1M
$1.50
at Mistral AI
Output / 1M
$7.50
at Mistral AI
Context window
262K
tokens
Max output
262K
tokens per reply
Hosts
2
3 listings
Released
Apr 2026
30 Apr 2026

Pricing

List prices at Mistral AI, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$1.50$7.50
Cached input
90% off input
$0.15

What it costs a month

At Mistral AI list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$891
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$666
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$1,116
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$5,400
Use your own numbers

Where to run it

2 hosts serve Mistral Medium 3.5. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Mistral AILab API
mistral-medium-3-5
$1.50 / $7.50
Retires 30 Apr 2027
mistral-medium-3-5
$1.50 / $7.50
Show 1 more listings: regions, snapshots and variants
Mistral AI
Alias
mistral-medium-3.5
$1.50 / $7.50

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
TextImages
Produces
Text
Effort settings the source lists
nonehigh
Features
Assistant prefill
Open weights
License: other
Runs in about 306.5 GB of GPU memory

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Mistral AI.

Smaller Mistral AI models

  • Magistral Small
    Mistral AI · on Mistral AI
    $0.15 / $0.60
    Has every key capability
    91% less
  • Mistral Vibe Cli Fast
    Mistral AI · on Mistral AI
    $0.15 / $0.60
    Has every key capability
    91% less
  • Zai Glm 5.3 Flash
    Mistral AI · on Mistral AI
    $0.15 / $0.50
    Has every key capability
    92% less

Other frontier labs

  • GPT-6.1 Terra
    OpenAI · on OpenAI
    $1.00 / $5.00
    Has every key capability
    33% less
  • Gemini 3.8 Flash
    Google · on Google AI Studio
    $0.75 / $3.75
    Has every key capability
    50% less
  • Grok Build 0.1
    xAI · on xAI
    $1.00 / $2.00
    Has every key capability
    58% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    93% less

Open-weight picks

  • Qwen4 Max
    Alibaba Qwen · on Alibaba DashScope
    $1.50 / $4.50
    Has every key capability
    25% less
  • Inkling
    Independent · on DeepInfra
    $0.95 / $4.05
    Has every key capability
    43% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    77% less
  • DeepSeek Flash
    DeepSeek · on DeepSeek
    $0.30 / $1.20
    Has every key capability
    83% less

Model line

Newer and older versions, at each lab's own price.

  1. UndatedMistral Medium 2604Latest$1.50 / $7.50
  2. UndatedMistral Medium 2505$0.40 / $2.00
  3. Apr 2026Mistral Medium 3.5This page$1.50 / $7.50
  4. May 2025Mistral Medium 3$1.50 / $7.50

Call it

Using Mistral AI's API.

Python
from litellm import completion

response = completion(
    model="mistral/mistral-medium-3-5",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Mistral Medium 3.5 cost?

On Mistral AI, Mistral Medium 3.5 costs $1.50 per million input tokens and $7.50 per million output tokens. Cached input costs $0.15 per million tokens.

Which APIs offer Mistral Medium 3.5?

Mistral Medium 3.5 runs on 2 APIs: Mistral AI and Azure AI Foundry.

What is Mistral Medium 3.5's context window?

262K tokens, with up to 262K tokens in one reply.

Is Mistral Medium 3.5 being retired?

Mistral AI (some versions only) retires it on 30 Apr 2027. 2 hosts keep serving it.

Is there a newer version of Mistral Medium 3.5?

Yes. Mistral Medium 2604 is the newest model in this line.

What are cheaper alternatives to Mistral Medium 3.5?

Models that cost less and keep most of what it can do: Magistral Small on Mistral AI (91% less), GPT-6.1 Terra on OpenAI (33% less), Gemini 3.8 Flash on Google AI Studio (50% less) and Qwen4 Max on Alibaba DashScope (25% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Mistral Medium 3.5 released?

Mistral AI released it on 30 Apr 2026.

What is Mistral Medium 3.5's API model id?

Use mistral-medium-3-5 with Mistral AI.

Sources

Raw catalog entry for mistral-medium-3-5 on Mistral AI
{
  "deprecation_date": "2027-04-30",
  "cache_read_input_token_cost": 1.5e-7,
  "input_cost_per_token": 0.0000015,
  "litellm_provider": "mistral",
  "max_input_tokens": 262144,
  "max_output_tokens": 262144,
  "mode": "chat",
  "output_cost_per_token": 0.0000075,
  "reasoning_effort_levels": [
    "none",
    "high"
  ],
  "source": "https://docs.mistral.ai/models/model-cards/mistral-medium-3-5-26-04",
  "supports_assistant_prefill": true,
  "supports_function_calling": true,
  "supports_reasoning": true,
  "supports_response_schema": true,
  "supports_tool_choice": true,
  "supports_vision": true
}