Coding
Directional only- BTL-4
- 78.4
- Claude Opus 5
- 89.5
- Weighted basis
- 1 vs 2 rows
- Reading
- Directional only
Model comparison
BTL-4 vs Claude Opus 5
Updated August 11, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
1 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
Claude Opus 5
Claude Opus 5 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
No shared weighted benchmark basis supports a winner.
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.
2 BTL-4 only
1 Shared
62 Claude Opus 5 only
- Shared results
- 1
- BTL-4 only
- 2
- Claude Opus 5 only
- 62
- Like-for-like categories
- 0 / 8
1 category uses 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.
Agentic
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- 90.8
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Reasoning
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- 90.4
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Knowledge
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- 64.7
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Math
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- 66.9
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Instruction following
Not comparable- BTL-4
- Not measured
- Claude Opus 5
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | BTL-4 | Claude Opus 5 | Weighted basis | Reading |
|---|---|---|---|---|
| Coding | 78.4 | 89.5 | Directional only1 vs 2 rows | Directional only |
| Agentic | Not measured | 90.8 | Not comparable0 vs 1 rows | Not comparable |
| Reasoning | Not measured | 90.4 | Not comparable0 vs 1 rows | Not comparable |
| Knowledge | Not measured | 64.7 | Not comparable0 vs 1 rows | Not comparable |
| Math | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 66.9 | Not comparable0 vs 1 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.
SWE-bench Verified
Coding
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
- BTL-4
- Self-hosted; infrastructure cost varies
- Fits in one request
- Claude Opus 5
- $0.0175
- Fits in one request
BTL-4 has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- BTL-4
- Self-hosted; infrastructure cost varies
- Fits in one request
- Claude Opus 5
- $0.325
- Fits in one request
BTL-4 has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- BTL-4
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
- Claude Opus 5
- $0.45
- Fits in one request
BTL-4 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.
SpecificationBTL-4Claude Opus 5
Context window
Maximum documented context; output-token limits may be lower.
BTL-4
Claude Opus 5
API model ID
BTL-4
Not sourced
Claude Opus 5
claude-opus-5
Anthropic model overviewCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
BTL-4
No comparable hosted API rate
Bad Theory Labs BTL-4 model cardClaude Opus 5
$0.5 per 1M cached input tokens
Claude API pricingDocumented inputs
BTL-4
Not sourced
Claude Opus 5
text, image
Anthropic model overviewDocumented outputs
BTL-4
Not sourced
Claude Opus 5
Provider availability
BTL-4
Not sourced
Claude Opus 5
Generally Available · Claude API
Anthropic model overviewReasoning profile
BTL-4
Reasoning
Claude Opus 5
Reasoning
Weight access
BTL-4
Open Weight
Claude Opus 5
Proprietary
License
BTL-4
Open Weight
Claude Opus 5
Proprietary
Release date
BTL-4
2026-08-05
Claude Opus 5
2026-07-24
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
- Claude Opus 5 has the larger documented window (1M).
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 evidence65 rows
Agentic
BFCL v4
BTL-473.5%SourceClaude Opus 5—Not directly comparable
FrontierBench v0.1
BTL-4—Claude Opus 543.3%SourceNot directly comparable
BrowseComp
BTL-4—Claude Opus 590.8%SourceNot directly comparable
HLE w/ tools
BTL-4—Claude Opus 564.7%SourceNot directly comparable
DeepSearchQA
BTL-4—Claude Opus 595.0%SourceNot directly comparable
DRACO
BTL-4—Claude Opus 588.6%SourceNot directly comparable
BrowseComp (10-agent, prerelease)
BTL-4—Claude Opus 593.6%SourceNot directly comparable
OSWorld 2.0
BTL-4—Claude Opus 570.6%SourceNot directly comparable
MCP Atlas
BTL-4—Claude Opus 585.8%SourceNot directly comparable
MCP-Atlas claim coverage
BTL-4—Claude Opus 589.1%SourceNot directly comparable
LAB all-pass (Anthropic harness)
BTL-4—Claude Opus 523.58%SourceNot directly comparable
LAB criterion-pass (Anthropic harness)
BTL-4—Claude Opus 593.74%SourceNot directly comparable
LAB all-pass (Harvey held-out)
BTL-4—Claude Opus 511.7%SourceNot directly comparable
LAB criterion-pass (Harvey held-out)
BTL-4—Claude Opus 594.1%SourceNot directly comparable
Toolathlon-Verified
BTL-4—Claude Opus 580.6%SourceNot directly comparable
Toolathlon Verified Pass@3
BTL-4—Claude Opus 587.0%SourceNot directly comparable
Toolathlon Verified Pass³
BTL-4—Claude Opus 573.1%SourceNot directly comparable
Toolathlon Verified avg. turns
BTL-4—Claude Opus 523.5 turnsSourceNot directly comparable
AutomationBench
BTL-4—Claude Opus 526.0%SourceNot directly comparable
Coding
SWE-bench Verified
BTL-478.4%SourceClaude Opus 596%SourceClaude Opus 5 leads this result
LiveCodeBench v6
BTL-466.1%SourceClaude Opus 5—Not directly comparable
SWE-bench Pro
BTL-4—Claude Opus 579.2%SourceNot directly comparable
SWE Multilingual
BTL-4—Claude Opus 589.5%SourceNot directly comparable
SWE Multimodal
BTL-4—Claude Opus 559.4%SourceNot directly comparable
deepSwe
BTL-4—Claude Opus 568.8%SourceNot directly comparable
FrontierCode 1.1 Main
BTL-4—Claude Opus 553.4%SourceNot directly comparable
FrontierCode 1.1 Extended
BTL-4—Claude Opus 563.6%SourceNot directly comparable
ProgramBench (episode 1)
BTL-4—Claude Opus 583.0%SourceNot directly comparable
ProgramBench
BTL-4—Claude Opus 593.0%SourceNot directly comparable
cursorBench32
BTL-4—Claude Opus 570.0%SourceNot directly comparable
VulcanBench v3
BTL-4—Claude Opus 587.0%SourceNot directly comparable
Reasoning
Knowledge
HLE
BTL-4—Claude Opus 564.7%SourceNot directly comparable
HLE w/o tools
BTL-4—Claude Opus 556.3%SourceNot directly comparable
HealthBench (raw)
BTL-4—Claude Opus 567.1%SourceNot directly comparable
HealthBench (length-adjusted)
BTL-4—Claude Opus 557.8%SourceNot directly comparable
HealthBench Professional
BTL-4—Claude Opus 559.8%SourceNot directly comparable
HealthBench Professional (raw)
BTL-4—Claude Opus 573.4%SourceNot directly comparable
BioMysteryBench (human-solvable)
BTL-4—Claude Opus 590.1%SourceNot directly comparable
BioMysteryBench (human-difficult)
BTL-4—Claude Opus 549.4%SourceNot directly comparable
SpatialBench Verified
BTL-4—Claude Opus 572.5%SourceNot directly comparable
SingleCellBench
BTL-4—Claude Opus 560.6%SourceNot directly comparable
ProteinGym Hard
BTL-4—Claude Opus 547.7%SourceNot directly comparable
Protein Design
BTL-4—Claude Opus 542.5%SourceNot directly comparable
Organic chemistry V2
BTL-4—Claude Opus 561.6%SourceNot directly comparable
Protocols (troubleshooting)
BTL-4—Claude Opus 561.1%SourceNot directly comparable
Protocols (understanding)
BTL-4—Claude Opus 578.4%SourceNot directly comparable
Math
IMO 2026
BTL-4—Claude Opus 542/42SourceNot directly comparable
RiemannBench (no tools)
BTL-4—Claude Opus 560.0%SourceNot directly comparable
RiemannBench (tools)
BTL-4—Claude Opus 579.0%SourceNot directly comparable
ArXivMath Jun. 2026 (no tools)
BTL-4—Claude Opus 590.8%SourceNot directly comparable
ArXivMath Jun. 2026 (tools)
BTL-4—Claude Opus 591.3%SourceNot directly comparable
Multilingual
Multimodal
Chartography (no tools)
BTL-4—Claude Opus 529.6%SourceNot directly comparable
Chartography (tools)
BTL-4—Claude Opus 583.0%SourceNot directly comparable
BenchCAD Vision2Code (no tools)
BTL-4—Claude Opus 50.366SourceNot directly comparable
BenchCAD Vision2Code (tools)
BTL-4—Claude Opus 50.821SourceNot directly comparable
GDP.pdf (no tools)
BTL-4—Claude Opus 583.4%SourceNot directly comparable
GDP.pdf (tools)
BTL-4—Claude Opus 585.5%SourceNot directly comparable
OfficeQA
BTL-4—Claude Opus 578.1%SourceNot directly comparable
OfficeQA Pro
BTL-4—Claude Opus 566.9%SourceNot directly comparable
Frequently asked questions
Which is better, BTL-4 or Claude Opus 5?
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, BTL-4 or Claude Opus 5?
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, BTL-4 or Claude Opus 5?
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, BTL-4 or Claude Opus 5?
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, BTL-4 or Claude Opus 5?
Claude Opus 5 has the larger documented context window: 1M, compared with 262K.
Related comparisons
Last updated August 11, 2026
Watch BTL-4 vs Claude Opus 5
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.