impulse

Muse Spark 1.2

Muse Spark 1.2 is a reasoning model from Meta, designed for complex agentic tasks.

API id muse-spark-1.2Released 5 Aug 2026 (Aug 2026)
Input / 1M
$1.25
at Meta Model API
Output / 1M
$4.25
at Meta Model API
Context window
1M
tokens
Max output
131K
tokens per reply
Hosts
4
4 listings
Released
Aug 2026
5 Aug 2026

Pricing

List prices at Meta Model API, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$1.25$4.25
Cached input
88% off input
$0.15
Web search
$2.50 per 1,000 searches

What it costs a month

At Meta Model API list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$595
  • RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$489
  • Coding agent1,000 requests a day, 40K in / 2K out, cached$831
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$4,020
Use your own numbers

Where to run it

4 hosts serve Muse Spark 1.2. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
muse-spark-1.2
$1.25 / $4.25
muse-spark-1-2
$1.25 / $4.25
AIHubMixGateway
muse-spark-1.2
$1.38 / $4.67
+10% vs headline
muse-spark-1.2
$1.77 / $6.03
+42% 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
ImagesTextVideoPDFs
Produces
Text
Effort settings the source lists
minimalxhigh
Features
Parallel tool callsPDF inputPrompt cachingBuilt-in web search

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 Meta Model API.

Smaller Meta models

Other frontier labs

  • GPT-5.4 Mini
    OpenAI · on OpenAI
    $0.75 / $4.50
    Has every key capability
    16% less
  • Gemini 3.8 Flash
    Google · on Google AI Studio
    $0.75 / $3.75
    Has every key capability
    25% less
  • Grok Build 0.1
    xAI · on xAI
    $1.00 / $2.00
    Has every key capability
    38% less
  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    90% less

Open-weight picks

  • Inkling
    Independent · on DeepInfra
    $0.95 / $4.05
    Has every key capability
    14% less
  • Kimi K2.7 Code
    Moonshot AI · on CLF AI Gateway
    $0.38 / $1.60
    Has every key capability
    66% less
  • DeepSeek Flash
    DeepSeek · on DeepSeek
    $0.30 / $1.20
    Has every key capability
    74% less
  • Qwen3.7 Plus
    Alibaba Qwen · on Replicate
    $0.276 / $1.10
    Has every key capability
    76% less

Model line

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

  1. Sep 2026Muse Spark 1.3Latest$1.25 / $4.25
  2. Aug 2026Muse Spark 1.2This page$1.25 / $4.25
  3. Jul 2026Muse Spark 1.1$1.25 / $4.25

Call it

Using Meta Model API's API.

Python
from litellm import completion

response = completion(
    model="meta/muse-spark-1.2",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Muse Spark 1.2 cost?

On Meta Model API, Muse Spark 1.2 costs $1.25 per million input tokens and $4.25 per million output tokens. Cached input costs $0.15 per million tokens.

Which APIs offer Muse Spark 1.2?

Muse Spark 1.2 runs on 4 APIs: Meta Model API, EmpirioLabs AI, AIHubMix and Kyma API.

What is Muse Spark 1.2's context window?

1M tokens, with up to 131K tokens in one reply.

Is Muse Spark 1.2 being retired?

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

Is there a newer version of Muse Spark 1.2?

Yes. Muse Spark 1.3 is the newest model in this line, released on 2 Sep 2026.

What are cheaper alternatives to Muse Spark 1.2?

Models that cost less and keep most of what it can do: Muse Spark 1.3 Contributor on Meta Model API (94% less), GPT-5.4 Mini on OpenAI (16% less), Gemini 3.8 Flash on Google AI Studio (25% less) and Inkling on DeepInfra (14% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Muse Spark 1.2 released?

Meta released it on 5 Aug 2026.

What is Muse Spark 1.2's API model id?

Use muse-spark-1.2 with Meta Model API. Other hosts use their own ids: muse-spark-1-2 on EmpirioLabs AI.

Sources

Raw catalog entry for muse-spark-1.2 on Meta Model API
{
  "cache_read_input_token_cost": 1.5e-7,
  "input_cost_per_token": 0.00000125,
  "litellm_provider": "meta",
  "max_input_tokens": 1048576,
  "max_output_tokens": 131072,
  "mode": "chat",
  "output_cost_per_token": 0.00000425,
  "search_context_cost_per_query": {
    "search_context_size_high": 0.0025,
    "search_context_size_low": 0.0025,
    "search_context_size_medium": 0.0025
  },
  "source": "https://ai.developer.meta.com/docs/pricing-rate-limits",
  "supported_endpoints": [
    "/v1/chat/completions",
    "/v1/responses",
    "/v1/messages"
  ],
  "supported_modalities": [
    "text",
    "image",
    "video"
  ],
  "supported_output_modalities": [
    "text"
  ],
  "supports_function_calling": true,
  "supports_minimal_reasoning_effort": true,
  "supports_parallel_function_calling": true,
  "supports_pdf_input": true,
  "supports_prompt_caching": true,
  "supports_reasoning": true,
  "supports_response_schema": true,
  "supports_system_messages": true,
  "supports_tool_choice": true,
  "supports_vision": true,
  "supports_web_search": true,
  "supports_xhigh_reasoning_effort": true
}