Muse Spark 1.3 and Muse Spark 1.3 Contributor Are Now Live on APIMaster.ai

Muse Spark 1.3 and Muse Spark 1.3 Contributor are now available in the APIMaster.ai marketplace. The live APIMaster routes currently list Muse Spark 1.3 at $1.25 input / $4.25 output per million tokens, while Muse Spark 1.3 Contributor starts at only $0.10 input / $0.20 output per million tokens. Prices, route availability, and quotas can change, so check the marketplace card before sending production traffic.

Meta describes Muse Spark 1.3 as a reasoning model built for long-horizon agentic work, coding, and practical engineering workflows. APIMaster gives developers one OpenAI-compatible endpoint, transparent per-token billing, route comparison, and model verification tools for testing the provider route they are about to use.

What is Muse Spark 1.3?

Meta’s official introduction presents Muse Spark 1.3 as a more capable and more usable model for agentic and coding tasks. It is designed to sustain longer work with multiple workflows in a single thread, build context from messy sources, correct gaps in its plan, and produce a final deliverable instead of stopping after one short answer.

The model is also trained to collaborate more actively. When a request is ambiguous, it can ask a clarifying question. When it is stuck, it can ask the user for help. Before consequential actions, it is designed to confirm rather than silently proceed. For long tasks, it can adapt to a user's preferred interaction style, with frequent progress updates or quieter background work.

What changed in Muse Spark 1.3?

Meta reports several practical improvements over Muse Spark 1.2:

• Better performance on long-horizon agentic and coding tasks.
• Fewer unnecessary turns and less verbose output.
• Cleaner coding style in engineering workflows.
• About 20% fewer tool calls and 25% fewer tokens in Meta engineer comparisons.
• More reliable handling of detailed, long-form instructions and constraints.
• Better awareness of limitations, with fewer claims of success when a task has not actually been completed.
• Stronger resistance to adversarial inputs and prompt injection.
• Better judgment about when an action is irreversible or consequential.

These are capabilities reported by Meta. Your results will still depend on the prompt, tools, context, system instructions, and route. Run a representative pilot before moving an agent into production.

Muse Spark 1.3 versus Muse Spark 1.3 Contributor

Model	APIMaster input price	APIMaster output price	Best fit
muse-spark-1.3	$1.25/M tokens	$4.25/M tokens	Standard access for agentic coding and long-running workflows
muse-spark-1.3-contributor	$0.10/M tokens	$0.20/M tokens	Experiments, prototypes, learning, and cost-sensitive workloads

Both model IDs are available through the same APIMaster API. OpenRouter describes Contributor as the cost-efficient contributor tier for experimentation, learning, and early-stage agentic and coding workflows. The price difference alone does not establish a difference in reasoning quality. Compare each route on your own tasks and review its current availability and applicable terms before choosing it for production.

OpenRouter lists both IDs with a 1,048,576-token context window and text, image, audio, video, and file inputs, with text output. Its published parameter list includes reasoning, tools, tool choice, and structured outputs. These are upstream catalog specifications; supported input formats and effective limits should be checked on the route you use.

APIMaster route tests on September 13, 2026: both IDs passed ordinary chat, streaming with token usage, structured JSON output, multi-turn context, and an automatic tool-call round trip. The current Meta route accepts only tool_choice: "auto"; named function choices, required, and none were rejected upstream. Use automatic tool selection for this route.

The table shows listed channel token prices. Your account's billing group multiplier may affect the final charge; check your effective rate and usage log. Reasoning tokens also contribute to output usage.

Why use Muse Spark through APIMaster.ai?

APIMaster is an API gateway and model marketplace for developers who want a practical way to compare models and providers:

• OpenAI-compatible API: use https://apimaster.ai/v1 with common SDKs and tools.
• Transparent token billing: see route prices and usage rather than paying for an opaque bundle.
• Live route comparison: review available channels, price, latency, and status in the marketplace.
• Model verification: use the AI model tester to check the behavior of a selected route when a completed fingerprint record is available.
• One API key: switch between Muse Spark, GPT, Claude, DeepSeek, Gemini, and other models without redesigning your integration.
• Developer tooling: connect the endpoint to agents, coding tools, Open WebUI, LangChain, LiteLLM, or your own application.

An available marketplace card means the route is listed and can change with upstream capacity. It is not a promise that every route supports every parameter. Confirm the endpoint, model ID, tool behavior, context needs, and error handling in your own test.

Quick start

1. Create an APIMaster account and add wallet credit if needed.
2. Open the model marketplace and select muse-spark-1.3 or muse-spark-1.3-contributor.
3. Create an API key and set your base URL to https://apimaster.ai/v1.
4. Copy the exact model ID from the marketplace and run a small representative request.

curl https://apimaster.ai/v1/chat/completions \
  -H "Authorization: Bearer $APIMASTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "muse-spark-1.3-contributor",
    "messages": [{"role": "user", "content": "Explain this code change and list the tests I should run."}],
    "max_tokens": 800
  }'

Start with Contributor for prompt development and low-cost evaluation. Compare the same task on standard Muse Spark 1.3 before choosing a route for a longer engineering workflow. Track output quality, tool calls, latency, and total token cost rather than judging a model from one chat response.

Try Muse Spark 1.3 on APIMaster.ai

Muse Spark 1.3 brings Meta’s latest improvements for agentic and coding workflows to an API that developers can use alongside their existing model stack. With Contributor available from $0.10 per million input tokens, it is easy to test the family at low cost before scaling. Open the APIMaster marketplace and start building.

FAQ

How much does Muse Spark 1.3 Contributor cost on APIMaster.ai?

As of September 13, 2026, the listed route costs $0.10 per million input tokens and $0.20 per million output tokens. Check the live marketplace for current prices and availability.

Which model IDs should I use?

Use muse-spark-1.3 or muse-spark-1.3-contributor with the APIMaster base URL https://apimaster.ai/v1.

Is the Contributor price evidence that it is a weaker model?

No. Contributor is listed as a cost-efficient tier, and price alone does not establish a quality difference. Evaluate the routes on the same prompts, tools, and acceptance criteria.

Does a listed model mean it has passed fingerprint verification?

No. Marketplace availability and fingerprint verification are separate. Check for a completed detection record for the specific model and route; this launch announcement does not claim Muse fingerprint verification.

Sources

• Meta AI Research — Introducing Muse Spark 1.3
• OpenRouter — Muse Spark 1.3
• OpenRouter — Muse Spark 1.3 Contributor
• APIMaster.ai model marketplace