impulse

Mistral Nemo

Open weights

A 12B parameter model with a 128k token context length built by Mistral in collaboration with NVIDIA.

API id mistral-nemoReleased 19 Jul 2024 (Jul 2024)
Input / 1M
$0.04
at Novita AI
Output / 1M
$0.17
at Novita AI
Context window
60K
tokens
Max output
16K
tokens per reply
Hosts
3
4 listings
Released
Jul 2024
19 Jul 2024

Pricing

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

TierInput / 1MOutput / 1M
Standard$0.04$0.17

What it costs a month

At Novita AI list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$24.75
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$25.32
  • Coding agent1,000 requests a day, 40K in / 2K out$58.20
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$137
Use your own numbers

Where to run it

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

HostIn / out per 1M
Novita AIInference
mistral-nemo
$0.04 / $0.17
mistral-nemo
$0.15 / $0.15
+107% vs headline
mistral-nemo@latest
$0.15 / $0.15
+107% vs headline
Show 1 more listings: regions, snapshots and variants
Vertex AI
Alias
mistral-nemo@2407
$3.00 / $3.00
+4038% vs headline

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
Text
Produces
Text
Open weights
License: Apache-2.0
Runs in about 29.4 GB of GPU memory
Runtimes: Ollama, llama.cpp, vLLM

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 Novita AI.

Smaller Mistral AI models

No cheaper Mistral AI model keeps its key capabilities.

Other frontier labs

No cheaper match from another frontier lab.

Open-weight picks

  • Command R7B
    Cohere · on Cohere Chat
    $0.0375 / $0.15
    Has every key capability
    9% less
  • Llama 3.2 3B
    Meta · on LlamaGate
    $0.04 / $0.08
    Has every key capability
    31% less
  • Qwen3.7 Flash Variant1
    Alibaba Qwen · on EmpirioLabs AI
    $0.028 / $0.11
    Has every key capability
    33% less
  • Gemma3 4B
    Google · on LlamaGate
    $0.03 / $0.08
    Has every key capability
    41% less

Benchmarks

Scores published on the model card.

  • BBH29.7%
  • GPQA5.4%
  • IFEval63.8%
  • MATH Lvl 512.7%
  • MMLU-Pro28.0%
  • MUSR8.5%
  • Open LLM Average24.7%

Call it

Using Novita AI's API.

Python
from litellm import completion

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

Quick answers

How much does Mistral Nemo cost?

On Novita AI, Mistral Nemo costs $0.04 per million input tokens and $0.17 per million output tokens.

Which APIs offer Mistral Nemo?

Mistral Nemo runs on 3 APIs: Novita AI, Azure AI Foundry and Vertex AI.

What is Mistral Nemo's context window?

60K tokens, with up to 16K tokens in one reply.

Is Mistral Nemo being retired?

No retirement date is scheduled for it on any API we track.

What are cheaper alternatives to Mistral Nemo?

Models that cost less and keep most of what it can do: Command R7B on Cohere Chat (9% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Mistral Nemo released?

Mistral AI released it on 19 Jul 2024.

What is Mistral Nemo's API model id?

Use mistral-nemo with Novita AI. Other hosts use their own ids: mistral-nemo@latest on Vertex AI.

Sources

Raw catalog entry for mistral-nemo on Novita AI
{
  "litellm_provider": "novita",
  "mode": "chat",
  "input_cost_per_token": 4e-8,
  "output_cost_per_token": 1.7e-7,
  "max_input_tokens": 60288,
  "max_output_tokens": 16000,
  "supports_tool_choice": true,
  "supports_system_messages": true,
  "supports_response_schema": true
}