Coding work
Code generation, repair, and software-engineering tasks
GPT-5.6 Luna
GPT-5.6 Luna leads on the public coding lane, 64.5 to 53.4, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
GPT-5.6 Luna vs Muse Spark
Updated September 24, 2026. Rank says GPT-5.6 Luna is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Share or export
Decision reading
GPT-5.6 Luna has the higher public score estimate, 65.6 versus 60.81, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 11 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 11
- GPT-5.6 Luna only
- 18
- Muse Spark only
- 16
- Like-for-like categories
- 3 / 8
Supported: GPT-5.6 Luna and Muse SparkHow the comparison works
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Agentic work
Tool use, computer use, and multi-step task completion
GPT-5.6 Luna
GPT-5.6 Luna leads on the public agentic lane, 55.3 to 53, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limitedLong documents
Prompts that approach the documented context limit
GPT-5.6 Luna
GPT-5.6 Luna has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-ratesCache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-ratesRepository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
64.5GPT-5.6 Luna53.4Muse Spark
Like-for-like · BenchAlign v5.7
GPT-5.6 Luna leads the like-for-like coding row, although the 90% intervals overlap.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
18 GPT-5.6 Luna only
11 Shared
16 Muse Spark only
1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
FrontierMath v2 (Tier 4)Math
Normalized gap 43.9GPT-5.6 Luna:58.500%Muse Spark:14.600%FrontierMath v2 (Tiers 1-3)Math
Normalized gap 39.6GPT-5.6 Luna:78.600%Muse Spark:39.000%ARC-AGI-2Reasoning
Normalized gap 17.0GPT-5.6 Luna:59.5%Muse Spark:42.5%SWE-bench ProCoding
Normalized gap 10.3GPT-5.6 Luna:62.7%Muse Spark:52.4%MMMU-ProMultimodal
Normalized gap 2.0GPT-5.6 Luna:78.4%Muse Spark:80.4%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.7 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Agentic
Like-for-like- GPT-5.6 Luna
- 55.3
- Supported · #28/105
- Muse Spark
- 53.0
- Supported · #33/105
- Basis
- BenchAlign v5.7 lane · 9 vs 5 public rows
- Reading
- GPT-5.6 Luna leads · intervals overlap
Coding
Like-for-like- GPT-5.6 Luna
- 64.5
- Supported · #9/135
- Muse Spark
- 53.4
- Supported · #32/135
- Basis
- BenchAlign v5.7 lane · 7 vs 5 public rows
- Reading
- GPT-5.6 Luna leads · intervals overlap
Knowledge
Like-for-like- GPT-5.6 Luna
- 64.6
- Supported · #22/158
- Muse Spark
- 60.8
- Supported · #33/158
- Basis
- BenchAlign v5.7 lane · 6 vs 7 public rows
- Reading
- GPT-5.6 Luna leads · intervals overlap
Multimodal
Directional only- GPT-5.6 Luna
- 67.1
- #22/50
- Muse Spark
- 77.5
- #14/50
- Basis
- Provisional lane · 1 vs 2 weighted rows
- Reading
- Directional only
Reasoning
Not comparable- GPT-5.6 Luna
- 54.7
- #18/19
- Muse Spark
- 53.3
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 2 vs 1 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-5.6 Luna
- Not ranked
- Muse Spark
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- GPT-5.6 Luna
- Not ranked
- Muse Spark
- 91.9
- #8/124
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- GPT-5.6 Luna
- 94.2
- Unranked · 3 rankable rows
- Muse Spark
- 55.1
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 2 vs 2 weighted rows
- Reading
- Not comparable
| Category | GPT-5.6 Luna | Muse Spark | Basis | Reading |
|---|---|---|---|---|
| Agentic | 55.3Supported · #28/105 | 53.0Supported · #33/105 | Like-for-likeBenchAlign v5.7 lane · 9 vs 5 public rows | GPT-5.6 Luna leads · intervals overlap |
| Coding | 64.5Supported · #9/135 | 53.4Supported · #32/135 | Like-for-likeBenchAlign v5.7 lane · 7 vs 5 public rows | GPT-5.6 Luna leads · intervals overlap |
| Knowledge | 64.6Supported · #22/158 | 60.8Supported · #33/158 | Like-for-likeBenchAlign v5.7 lane · 6 vs 7 public rows | GPT-5.6 Luna leads · intervals overlap |
| Multimodal | 67.1#22/50 | 77.5#14/50 | Directional onlyProvisional lane · 1 vs 2 weighted rows | Directional only |
| Reasoning | 54.7#18/19 | 53.3Unranked · 3 rankable rows | Not comparableProvisional lane · 2 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | 91.9#8/124 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | 94.2Unranked · 3 rankable rows | 55.1Unranked · 2 rankable rows | Not comparableProvisional lane · 2 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- GPT-5.6 Luna
- $0.0008
- Fits in one request
- Muse Spark
- API rate not published
- Fits in one request
Muse Spark has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- GPT-5.6 Luna
- $0.0136
- Fits in one request
- Muse Spark
- API rate not published
- Fits in one request
Muse Spark has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-5.6 Luna
- $0.02
- Fits in one request
- Muse Spark
- API rate not published
- Fits in one request
- Cached-input rate unavailable
Muse Spark has no comparable published API token rate.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGPT-5.6 LunaMuse Spark
Context window
Maximum documented context; output-token limits may be lower.
GPT-5.6 Luna
1.05M
OpenAI model catalogMuse Spark
262K
API model ID
GPT-5.6 Luna
gpt-5.6-luna
OpenAI model catalogMuse Spark
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-5.6 Luna
$0.02 per 1M cached input tokens
OpenAI pricingMuse Spark
No comparable hosted API rate
Documented inputs
GPT-5.6 Luna
text, image
OpenAI model catalogMuse Spark
Not sourced
Documented outputs
GPT-5.6 Luna
Muse Spark
Not sourced
Provider availability
GPT-5.6 Luna
Generally Available · OpenAI Responses API
OpenAI model catalogMuse Spark
Not sourced
Reasoning profile
GPT-5.6 Luna
Reasoning
Muse Spark
Reasoning
Weight access
GPT-5.6 Luna
Proprietary
Muse Spark
Proprietary
License
GPT-5.6 Luna
Proprietary
Muse Spark
Proprietary
Release date
GPT-5.6 Luna
2026-07-09
Muse Spark
2026-04-08
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- GPT-5.6 Luna has the higher public score estimate, 65.6 versus 60.81, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- GPT-5.6 Luna has the larger documented window (1.05M).
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, GPT-5.6 Luna or Muse Spark?
GPT-5.6 Luna has the higher public score estimate, 65.6 versus 60.81, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, GPT-5.6 Luna or Muse Spark?
GPT-5.6 Luna leads the public coding lane, 64.5 to 53.4, with Supported evidence for both models, although the 90% intervals overlap.
Which is better for agentic tasks, GPT-5.6 Luna or Muse Spark?
GPT-5.6 Luna leads the public agentic tasks lane, 55.3 to 53, with Supported evidence for both models, although the 90% intervals overlap.
Which costs less, GPT-5.6 Luna or Muse Spark?
Both models do not have comparable published API token rates, so this page does not name a universal price winner.
Which has the larger context window, GPT-5.6 Luna or Muse Spark?
GPT-5.6 Luna has the larger documented context window: 1.05M, compared with 262K.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence45 rows
Agentic
Terminal-Bench 3.0
GPT-5.6 Luna14.3%SourceMuse Spark—Not directly comparable
Terminal-Bench 2.1
GPT-5.6 Luna84.7%SourceMuse Spark—Not directly comparable
BrowseComp
GPT-5.6 Luna83.3%SourceMuse Spark—Not directly comparable
OSWorld 2.0
GPT-5.6 Luna45.6%SourceMuse Spark—Not directly comparable
CyberGym
GPT-5.6 Luna77.9%SourceMuse Spark43.5%SourceGPT-5.6 Luna leads this result
ExploitGym
GPT-5.6 Luna12.4%SourceMuse Spark—Not directly comparable
Toolathlon
GPT-5.6 Luna53.4%SourceMuse Spark—Not directly comparable
Terminal-Bench 2.1 (Vals)
GPT-5.6 Luna79.0%SourceMuse Spark—Not directly comparable
ApprenticeBench
GPT-5.6 Luna7%SourceMuse Spark—Not directly comparable
Terminal-Bench 2.0
GPT-5.6 Luna—Muse Spark59%SourceNot directly comparable
τ²-bench results
GPT-5.6 Luna—Muse Spark91.5%SourceNot directly comparable
DeepSearchQA
GPT-5.6 Luna—Muse Spark74.8%SourceNot directly comparable
Claw-Eval
GPT-5.6 Luna—Muse Spark63.8%SourceNot directly comparable
Coding
SWE-bench Pro
GPT-5.6 Luna62.7%SourceMuse Spark52.4%SourceGPT-5.6 Luna leads this result
Terminal-Bench 2.1
GPT-5.6 Luna84.7%SourceMuse Spark—Not directly comparable
DeepSWE
GPT-5.6 Luna67.2%SourceMuse Spark—Not directly comparable
FrontierCode 1.1 Extended
GPT-5.6 Luna55.1%SourceMuse Spark—Not directly comparable
cursorBench32
GPT-5.6 Luna61.1%SourceMuse Spark—Not directly comparable
VulcanBench v3
GPT-5.6 Luna85.5%SourceMuse Spark—Not directly comparable
SWE-bench (Vals)
GPT-5.6 Luna93.0%SourceMuse Spark74.4%SourceGPT-5.6 Luna leads this result
SWE-bench Verified
GPT-5.6 Luna—Muse Spark77.4%SourceNot directly comparable
LiveCodeBench Pro
GPT-5.6 Luna—Muse Spark80.0%SourceNot directly comparable
Vibe Code Bench
GPT-5.6 Luna—Muse Spark19.67%SourceNot directly comparable
Reasoning
Multimodal
MMMU-Pro
GPT-5.6 Luna78.4%SourceMuse Spark80.4%SourceMuse Spark leads this result
MMMU-Pro w/ Python
GPT-5.6 Luna79.5%SourceMuse Spark—Not directly comparable
CharXiv
GPT-5.6 Luna—Muse Spark86.4%SourceNot directly comparable
ERQA
GPT-5.6 Luna—Muse Spark64.7%SourceNot directly comparable
SimpleVQA
GPT-5.6 Luna—Muse Spark71.3%SourceNot directly comparable
ScreenSpot Pro
GPT-5.6 Luna—Muse Spark84.1%SourceNot directly comparable
ZeroBench
GPT-5.6 Luna—Muse Spark33.0%SourceNot directly comparable
MedXpertQA (MM)
GPT-5.6 Luna—Muse Spark78.4%SourceNot directly comparable
Knowledge
GPQA
GPT-5.6 Luna92.3%SourceMuse Spark—Not directly comparable
GPQA-D
GPT-5.6 Luna92.3%SourceMuse Spark89.5%SourceGPT-5.6 Luna leads this result
HealthBench Professional
GPT-5.6 Luna55.7%SourceMuse Spark—Not directly comparable
HealthBench Hard
GPT-5.6 Luna32.0%SourceMuse Spark42.8%SourceMuse Spark leads this result
GPQA Diamond (Vals)
GPT-5.6 Luna91.7%SourceMuse Spark89.6%SourceGPT-5.6 Luna leads this result
MMLU-Pro (Vals)
GPT-5.6 Luna86.0%SourceMuse Spark87.3%SourceMuse Spark leads this result
HLE
GPT-5.6 Luna—Muse Spark50.4%SourceNot directly comparable
HLE w/o tools
GPT-5.6 Luna—Muse Spark42.8%SourceNot directly comparable
MedXpertQA (Text)
GPT-5.6 Luna—Muse Spark52.6%SourceNot directly comparable
Math
45 public results · 11 shared
Watch GPT-5.6 Luna vs Muse Spark
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.
Last updated September 24, 2026