Muse Spark 1.3
Muse Spark 1.3Alternatives?
CheaperGemini 3.8 Flashunstatedwins 45%-25%*$0.75 / $3.75Sampling parameters are ignored+1
BetterGPT-6.1 Solmaxwins 57%+100%$0.72Remove unsupported parameters+5
OpenKimi K3unstatedsame+200%$3 / $15Sampling parameters are fixed+3
CheaperDeepSeek V4.1 Flash 1017unstatedOpenwins 45%-74%$0.30 / $1.20No PDF input+5
CheaperStep 3.7 FlashunstatedOpenwins 45%-78%$0.20 / $1.15Context: 1M → 256K+1Gemini 3.8 Flashunstatedwins 45%-25%*$0.75 / $3.75Sampling parameters are ignored+1DeepSeek V4.1 Flash 1017unstatedOpenwins 45%-74%$0.30 / $1.20No PDF input+5Step 3.7 FlashunstatedOpenwins 45%-78%$0.20 / $1.15Context: 1M → 256K+1Gemini 3.6 Flashunstatedwins 44%-25%*$0.75 / $3.75
Qwen3.5 2BunstatedOpenwins 43%-98%$0.02 / $0.10Context: 1M → 262K+1GPT-6.1 Solmaxwins 57%+100%$0.72Remove unsupported parameters+5GPT-6 Astralowwins 56%+900%$0.82Monitoring can return HTTP 403+2GPT-6 Solmaxwins 56%+100%$1.04
Claude Opus 5.5highwins 57%+300%$1.82No forced tool use+5Claude Fable 5.1lowwins 55%+900%$2.37Replace forced tool_choice+5Kimi K3unstatedOpensame+200%$3 / $15Sampling parameters are fixed+3Muse Spark OpenunstatedOpensamesame$1.25 / $4.25Reasoning off leaves output unchanged · Self-hosted / llama.cpp+1DeepSeek V4.1 Flash 1017unstatedOpenwins 45%-74%$0.30 / $1.20No PDF input+5Step 3.7 FlashunstatedOpenwins 45%-78%$0.20 / $1.15Context: 1M → 256K+1
GLM 5.2unstatedOpenwins 44%+7%$1.40 / $4.40No image input+1Muse Spark 1.1unstatedsame—$1.25 / $4.25Only n=1 · Meta / Chat Completions+1Muse Spark OpenunstatedOpensame—$1.25 / $4.25Reasoning off leaves output unchanged · Self-hosted / llama.cpp+1Muse Wondermaxn/a—$4.52Only auto tool_choice is supported+5Muse Wonder Fulln/a——Only in Muse CodeMuse Glimmer 30BhighOpenn/a—$0.058Context: 1M → 131KMuse Spark 1.2xhighn/a—$0.97Only auto tool_choice is supported+1Muse Sparkthinkingn/a—$1.25 / $4.25Only n=1 · Meta / Chat Completions+1Llama 4 Mavericknon-thinkingOpenn/a—$0.038No PDF inputAll Meta models →History
4 versionsBenchmarks
in Chat & KnowledgeWatch out 6
Only auto tool_choice is supported›
tool_choice none, required or a named function returns HTTP 400; use auto and handle either a tool call or text. Source ↗
Search filters return HTTP 400 · Messages API›
allowed_domains, blocked_domains and max_uses return HTTP 400 on Messages search grounding; omit them and use the supported web_search tool. Source ↗
Manual budget does not set effort · Messages API›
Compatibility budget_tokens does not set reasoning effort; use output_config.effort for depth control. Source ↗
No Chat Completions search grounding · Chat Completions›
Search grounding is unavailable in Chat Completions; use Responses or Messages. Source ↗
Reasoning cannot be disabled›
Muse Wonder and Muse Spark do not support disabling reasoning with none; keep reasoning enabled. Source ↗
Private reasoning still costs tokens›
Private reasoning consumes the output budget and billed output tokens; budget for it even though it is absent from the response body. Source ↗
FAQ
How much does Muse Spark 1.3 cost?›
$1.25 per 1M input tokens and $4.25 per 1M output tokens; cached input $0.15 per 1M. A typical benchmark task costs about $1.37 at xhigh effort.
Is Muse Spark 1.3 free to use?›
There is no free API; the cheapest paid price is $1.25 per 1M input tokens.
How good is Muse Spark 1.3, and what is it best at?›
#17 of 334 in Chat & Knowledge, #30 of 132 in Coding, #26 of 133 in Agents & Tool Use, #17 of 195 in Writing & Language, #11 of 124 in Vision & Documents. Strongest at: Coding, Agents & work tasks, Reasoning, Science, Long context, Knowledge, Vision.
Is Muse Spark 1.3 better than GPT-5.6 Sol?›
On MathArena · BrokenArXiv · June 2026, Muse Spark 1.3 scores 53.2 vs 67.3 — about the same. It is 31% cheaper per task.
How can I use Muse Spark 1.3?›
Through the API of Meta.
What is the context window of Muse Spark 1.3?›
1M tokens, with up to 944K output tokens.
When was Muse Spark 1.3 released? Is there a newer version?›
Released 2026-09-02. It is the current version in its line.
Is Muse Spark 1.3 open source?›
No, Muse Spark 1.3 is a closed model available through an API. The closest open-weights option here is Kimi K3.
Can Muse Spark 1.3 read images?›
It accepts text and images, PDF files, video as input.
What is Muse Spark 1.3?›
Muse Spark 1.3 is a closed language model from Meta, released Sep 2026, the successor to Muse Spark 1.2.
What are the best alternatives to Muse Spark 1.3?›
Gemini 3.8 Flash (cheaper), GPT-6.1 Sol (better), Kimi K3 (open weights).