Reasoning
Like-for-like- dots3-note Preview
- 81.4
- GPT-5.6 Luna
- 59.5
- Weighted basis
- 1 vs 1 rows
- Reading
- dots3-note Preview leads
dots3-note Preview vs GPT-5.6 Luna
Updated August 14, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
4 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
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.
Long 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
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
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-rates
Cache-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-rates
Repository 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
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
27 dots3-note Preview only
4 Shared
19 GPT-5.6 Luna only
- Shared results
- 4
- dots3-note Preview only
- 27
- GPT-5.6 Luna only
- 19
- Like-for-like categories
- 2 / 8
2 categories use different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Multimodal
Like-for-like- dots3-note Preview
- 79.1
- GPT-5.6 Luna
- 78.4
- Weighted basis
- 1 vs 1 rows
- Reading
- dots3-note Preview leads
Agentic
Directional only- dots3-note Preview
- 83.3
- GPT-5.6 Luna
- 84.1
- Weighted basis
- 1 vs 2 rows
- Reading
- Directional only
Coding
Directional only- dots3-note Preview
- 71.7
- GPT-5.6 Luna
- 62.7
- Weighted basis
- 2 vs 1 rows
- Reading
- Directional only
Knowledge
Not comparable- dots3-note Preview
- 52.6
- GPT-5.6 Luna
- 92.3
- Weighted basis
- 1 vs 1 rows
- Reading
- Not comparable
Math
Not comparable- dots3-note Preview
- Not measured
- GPT-5.6 Luna
- 73.6
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Multilingual
Not comparable- dots3-note Preview
- Not measured
- GPT-5.6 Luna
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Instruction following
Not comparable- dots3-note Preview
- 85.1
- GPT-5.6 Luna
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
| Category | dots3-note Preview | GPT-5.6 Luna | Weighted basis | Reading |
|---|---|---|---|---|
| Reasoning | 81.4 | 59.5 | Like-for-like1 vs 1 rows | dots3-note Preview leads |
| Multimodal | 79.1 | 78.4 | Like-for-like1 vs 1 rows | dots3-note Preview leads |
| Agentic | 83.3 | 84.1 | Directional only1 vs 2 rows | Directional only |
| Coding | 71.7 | 62.7 | Directional only2 vs 1 rows | Directional only |
| Knowledge | 52.6 | 92.3 | Not comparable1 vs 1 rows | Not comparable |
| Math | Not measured | 73.6 | Not comparable0 vs 2 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Instruction following | 85.1 | Not measured | Not comparable2 vs 0 rows | Not comparable |
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.
ARC-AGI-2
Reasoning
dots3-note Preview: 81.4%GPT-5.6 Luna: 59.5%Normalized gap 21.9dots3-note Preview source GPT-5.6 Luna sourceSWE-bench Pro
Coding
dots3-note Preview: 61%GPT-5.6 Luna: 62.7%Normalized gap 1.7dots3-note Preview source GPT-5.6 Luna sourceMMMU-Pro
Multimodal
dots3-note Preview: 79.1%GPT-5.6 Luna: 78.4%Normalized gap 0.7dots3-note Preview source GPT-5.6 Luna source
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
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- GPT-5.6 Luna
- $0.0008
- Fits in one request
dots3-note Preview has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- GPT-5.6 Luna
- $0.0136
- Fits in one request
dots3-note Preview has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- dots3-note Preview
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
- GPT-5.6 Luna
- $0.02
- Fits in one request
dots3-note Preview has no comparable published API token rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
Specificationdots3-note PreviewGPT-5.6 Luna
Context window
Maximum documented context; output-token limits may be lower.
dots3-note Preview
GPT-5.6 Luna
1.05M
OpenAI model catalogAPI model ID
dots3-note Preview
Not sourced
GPT-5.6 Luna
gpt-5.6-luna
OpenAI model catalogCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
dots3-note Preview
No comparable hosted API rate
dots3-note Preview model cardGPT-5.6 Luna
$0.02 per 1M cached input tokens
OpenAI pricingDocumented inputs
dots3-note Preview
Not sourced
GPT-5.6 Luna
text, image
OpenAI model catalogDocumented outputs
dots3-note Preview
Not sourced
GPT-5.6 Luna
Provider availability
dots3-note Preview
Not sourced
GPT-5.6 Luna
Generally Available · OpenAI Responses API
OpenAI model catalogReasoning profile
dots3-note Preview
Reasoning
GPT-5.6 Luna
Reasoning
Weight access
dots3-note Preview
Open Weight
GPT-5.6 Luna
Proprietary
License
dots3-note Preview
Open Weight
GPT-5.6 Luna
Proprietary
Release date
dots3-note Preview
2026-08-14
GPT-5.6 Luna
2026-07-09
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
- At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
- 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.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence50 rows
Agentic
Claw-Eval
dots3-note Preview73.4%SourceGPT-5.6 Luna—Not directly comparable
Terminal-Bench 2.1
dots3-note Preview75.1%SourceGPT-5.6 Luna—Not directly comparable
Toolathlon-Verified
dots3-note Preview55.6%SourceGPT-5.6 Luna—Not directly comparable
skillsBench
dots3-note Preview52.8%SourceGPT-5.6 Luna—Not directly comparable
APEX-Agents
dots3-note Preview30.8%SourceGPT-5.6 Luna—Not directly comparable
BrowseComp
dots3-note Preview83.3%SourceGPT-5.6 Luna83.3%SourceTie
HLE w/ tools
dots3-note Preview52.6%SourceGPT-5.6 Luna—Not directly comparable
DeepSearchQA
dots3-note Preview92.1%SourceGPT-5.6 Luna—Not directly comparable
WideResearch
dots3-note Preview78.9%SourceGPT-5.6 Luna—Not directly comparable
Terminal-Bench 3.0
dots3-note Preview—GPT-5.6 Luna14.3%SourceNot directly comparable
Terminal-Bench 2.0
dots3-note Preview—GPT-5.6 Luna84.7%SourceNot directly comparable
OSWorld 2.0
dots3-note Preview—GPT-5.6 Luna45.6%SourceNot directly comparable
CyberGym
dots3-note Preview—GPT-5.6 Luna77.9%SourceNot directly comparable
ExploitGym
dots3-note Preview—GPT-5.6 Luna12.4%SourceNot directly comparable
Toolathlon
dots3-note Preview—GPT-5.6 Luna53.4%SourceNot directly comparable
Coding
Codeforces
dots3-note Preview3056.0SourceGPT-5.6 Luna—Not directly comparable
LiveCodeBench v6
dots3-note Preview91.5%SourceGPT-5.6 Luna—Not directly comparable
Terminal-Bench 2.1
dots3-note Preview75.1%SourceGPT-5.6 Luna—Not directly comparable
SWE-bench Verified
dots3-note Preview78.4%SourceGPT-5.6 Luna—Not directly comparable
SWE Multilingual
dots3-note Preview75.7%SourceGPT-5.6 Luna—Not directly comparable
SWE-bench Pro
dots3-note Preview61%SourceGPT-5.6 Luna62.7%SourceGPT-5.6 Luna leads this result
NL2Repo
dots3-note Preview49.8%SourceGPT-5.6 Luna—Not directly comparable
Terminal-Bench 2.0
dots3-note Preview—GPT-5.6 Luna84.7%SourceNot directly comparable
deepSwe
dots3-note Preview—GPT-5.6 Luna67.2%SourceNot directly comparable
FrontierCode 1.1 Extended
dots3-note Preview—GPT-5.6 Luna55.1%SourceNot directly comparable
cursorBench32
dots3-note Preview—GPT-5.6 Luna61.1%SourceNot directly comparable
Reasoning
Knowledge
HLE
dots3-note Preview52.6%SourceGPT-5.6 Luna—Not directly comparable
GPQA
dots3-note Preview—GPT-5.6 Luna92.3%SourceNot directly comparable
GPQA-D
dots3-note Preview—GPT-5.6 Luna92.3%SourceNot directly comparable
HealthBench Professional
dots3-note Preview—GPT-5.6 Luna55.7%SourceNot directly comparable
HealthBench Hard
dots3-note Preview—GPT-5.6 Luna32.0%SourceNot directly comparable
Math
IMOAnswerBench
dots3-note Preview90.9%SourceGPT-5.6 Luna—Not directly comparable
FrontierMath (legacy)
dots3-note Preview—GPT-5.6 Luna78.6%SourceNot directly comparable
FrontierMath v2 (Tiers 1-3)
dots3-note Preview—GPT-5.6 Luna78.600%SourceNot directly comparable
FrontierMath v2 (Tier 4)
dots3-note Preview—GPT-5.6 Luna58.500%SourceNot directly comparable
Multimodal
SimpleVQA
dots3-note Preview72.5%SourceGPT-5.6 Luna—Not directly comparable
MMMU-Pro
dots3-note Preview79.1%SourceGPT-5.6 Luna78.4%Sourcedots3-note Preview leads this result
MathVision
dots3-note Preview87.7%SourceGPT-5.6 Luna—Not directly comparable
ZeroBench
dots3-note Preview19.0%SourceGPT-5.6 Luna—Not directly comparable
CharXiv w/o tools
dots3-note Preview83.1%SourceGPT-5.6 Luna—Not directly comparable
GDP.pdf (no tools)
dots3-note Preview60.7%SourceGPT-5.6 Luna—Not directly comparable
PerceptionBench
dots3-note Preview53.4%SourceGPT-5.6 Luna—Not directly comparable
BabyVision
dots3-note Preview50.0%SourceGPT-5.6 Luna—Not directly comparable
MMVU
dots3-note Preview79.9%SourceGPT-5.6 Luna—Not directly comparable
VideoMMMU
dots3-note Preview86.8%SourceGPT-5.6 Luna—Not directly comparable
MMMU-Pro w/ Python
dots3-note Preview—GPT-5.6 Luna79.5%SourceNot directly comparable
Instruction following
Frequently asked questions
Which is better, dots3-note Preview or GPT-5.6 Luna?
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, dots3-note Preview or GPT-5.6 Luna?
The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, dots3-note Preview or GPT-5.6 Luna?
The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which costs less, dots3-note Preview or GPT-5.6 Luna?
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, dots3-note Preview or GPT-5.6 Luna?
GPT-5.6 Luna has the larger documented context window: 1.05M, compared with 512K.
Related comparisons
Last updated August 14, 2026
Watch dots3-note Preview vs GPT-5.6 Luna
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.