Math
Like-for-like- Claude Sonnet 4.6
- 26.4
- o3
- 14.5
- Weighted basis
- 2 vs 2 rows
- Reading
- Claude Sonnet 4.6 leads
Claude Sonnet 4.6 vs o3
Updated September 3, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Claude Sonnet 4.6 has the higher public score, 64.57 versus 46.83, and the 90% score intervals do not overlap.
2 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.
Chat turn cost
1K fresh input + 500 output tokens
o3
o3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Show secondary and unsupported calls
Repository review cost
50K fresh input + 3K output tokens
o3
o3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
Long documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Claude Sonnet 4.6 does not fit this workload in one request. o3 does not fit this workload in one request. Claude Sonnet 4.6 has no published cached-input rate, so cached tokens use its listed input rate. o3 has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
18 Claude Sonnet 4.6 only
2 Shared
- Shared results
- 2
- Claude Sonnet 4.6 only
- 18
- o3 only
- 0
- Like-for-like categories
- 1 / 8
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.
Agentic
Not comparable- Claude Sonnet 4.6
- 65.2
- o3
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Coding
Not comparable- Claude Sonnet 4.6
- 69.1
- o3
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Reasoning
Not comparable- Claude Sonnet 4.6
- Not measured
- o3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Knowledge
Not comparable- Claude Sonnet 4.6
- 66.0
- o3
- Not measured
- Weighted basis
- 4 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Sonnet 4.6
- Not measured
- o3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Sonnet 4.6
- 77.4
- o3
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Sonnet 4.6
- Not measured
- o3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Claude Sonnet 4.6 | o3 | Weighted basis | Reading |
|---|---|---|---|---|
| Math | 26.4 | 14.5 | Like-for-like2 vs 2 rows | Claude Sonnet 4.6 leads |
| Agentic | 65.2 | Not measured | Not comparable2 vs 0 rows | Not comparable |
| Coding | 69.1 | Not measured | Not comparable2 vs 0 rows | Not comparable |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Knowledge | 66.0 | Not measured | Not comparable4 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | 77.4 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Instruction following | Not measured | Not measured | Not comparable0 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.
FrontierMath v2 (Tiers 1-3)
Math
FrontierMath v2 (Tier 4)
Math
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
- Claude Sonnet 4.6
- $0.0105
- Fits in one request
- o3
- $0.006
- Fits in one request
o3 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Claude Sonnet 4.6
- $0.195
- Fits in one request
- o3
- $0.124
- Fits in one request
o3 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Claude Sonnet 4.6
- $0.81
- Does not fit in one request
- Cached input priced at the published list-input rate
- o3
- $0.52
- Does not fit in one request
- Cached input priced at the published list-input rate
Claude Sonnet 4.6 does not fit this workload in one request. o3 does not fit this workload in one request. Claude Sonnet 4.6 has no published cached-input rate, so cached tokens use its listed input rate. o3 has no published cached-input rate, so cached tokens use its listed input rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationClaude Sonnet 4.6o3
Context window
Maximum documented context; output-token limits may be lower.
Claude Sonnet 4.6
200K
o3
200K
API model ID
Claude Sonnet 4.6
Not sourced
o3
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Sonnet 4.6
Not published
o3
Not published
Documented inputs
Claude Sonnet 4.6
Not sourced
o3
Not sourced
Documented outputs
Claude Sonnet 4.6
Not sourced
o3
Not sourced
Provider availability
Claude Sonnet 4.6
Not sourced
o3
Not sourced
Reasoning profile
Claude Sonnet 4.6
Non-Reasoning
o3
Reasoning
Weight access
Claude Sonnet 4.6
Proprietary
o3
Proprietary
License
Claude Sonnet 4.6
Proprietary
o3
Proprietary
Release date
Claude Sonnet 4.6
2026-02-01
o3
2025-04-16
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
- Claude Sonnet 4.6 has the higher public score, 64.57 versus 46.83, and the 90% score intervals do not overlap.
- Workload cost
- Repository review: $0.195 vs $0.124. Cache-heavy agent loop: $0.81 vs $0.52.
- Context tradeoff
- Both models list 200K.
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 evidence20 rows
Agentic
Terminal-Bench 2.0
Claude Sonnet 4.659.1%Sourceo3—Not directly comparable
OSWorld-Verified
Claude Sonnet 4.672.1%Sourceo3—Not directly comparable
Claw-Eval
Claude Sonnet 4.667.8%Sourceo3—Not directly comparable
CyberGym
Claude Sonnet 4.665.2%Sourceo3—Not directly comparable
Gert Labs
Claude Sonnet 4.662.92%Sourceo3—Not directly comparable
OSWorld 2.0
Claude Sonnet 4.68.3%Sourceo3—Not directly comparable
JobBench
Claude Sonnet 4.636.9%Sourceo3—Not directly comparable
Coding
SWE-bench Verified
Claude Sonnet 4.679.6%Sourceo3—Not directly comparable
SWE-Rebench
Claude Sonnet 4.660.7%Sourceo3—Not directly comparable
React Native Evals
Claude Sonnet 4.680.6%Sourceo3—Not directly comparable
Vibe Code Bench
Claude Sonnet 4.651.48%Sourceo3—Not directly comparable
cursorBench31
Claude Sonnet 4.648.8%Sourceo3—Not directly comparable
FrontierCode 1.1 Main
Claude Sonnet 4.624.3%Sourceo3—Not directly comparable
Knowledge
Math
FrontierMath v2 (Tiers 1-3)
Shared sourceClaude Sonnet 4.632.400%o318.685%Claude Sonnet 4.6 leads this result
FrontierMath v2 (Tier 4)
Shared sourceClaude Sonnet 4.68.300%o32.083%Claude Sonnet 4.6 leads this result
Multimodal
CharXiv
Claude Sonnet 4.677.4%Sourceo3—Not directly comparable
Frequently asked questions
Which is better, Claude Sonnet 4.6 or o3?
Claude Sonnet 4.6 has the higher public score, 64.57 versus 46.83, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.
Which is better for coding, Claude Sonnet 4.6 or o3?
The published evidence does not provide a shared weighted coding basis for both models, so BenchLM does not name a coding winner.
Which is better for agentic tasks, Claude Sonnet 4.6 or o3?
The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.
Which costs less, Claude Sonnet 4.6 or o3?
For the stated presets, chat costs $0.0105 on Claude Sonnet 4.6 and $0.006 on o3; repository review costs $0.195 and $0.124; the cache-heavy agent loop costs $0.81 and $0.52. Claude Sonnet 4.6 does not fit this workload in one request. o3 does not fit this workload in one request. Claude Sonnet 4.6 has no published cached-input rate, so cached tokens use its listed input rate. o3 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Claude Sonnet 4.6 or o3?
Both models list the same context window, 200K.
Related comparisons
Last updated September 3, 2026
Watch Claude Sonnet 4.6 vs o3
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.