Agentic
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- 75.1
- Weighted basis
- 0 vs 3 rows
- Reading
- Not comparable
Claude Mythos Preview vs Claude Opus 4.7 (Adaptive)
Updated August 13, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
AN
Model BAnthropic
72.2/100
Estimated · Public rank #14
90% interval 62.3–82.0
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.
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.
Chat turn cost
1K fresh input + 500 output tokens
Claude Opus 4.7 (Adaptive)
Claude Opus 4.7 (Adaptive) has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Claude Opus 4.7 (Adaptive)
Claude Opus 4.7 (Adaptive) has the lower estimated token cost for this stated workload. Claude Mythos Preview has no published cached-input rate, so cached tokens use its listed input rate. Claude Opus 4.7 (Adaptive) has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback
Show secondary and unsupported calls
Repository review cost
50K fresh input + 3K output tokens
Claude Opus 4.7 (Adaptive)
Claude Opus 4.7 (Adaptive) 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
Not enough matched evidence
A complete context comparison is not sourced.
Confidence: limited
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
1 Claude Mythos Preview only
1 Shared
20 Claude Opus 4.7 (Adaptive) only
- Shared results
- 1
- Claude Mythos Preview only
- 1
- Claude Opus 4.7 (Adaptive) only
- 20
- Like-for-like categories
- 0 / 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.
Coding
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- 78.6
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Reasoning
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- 75.8
- Weighted basis
- 0 vs 1 rows
- Reading
- Not comparable
Knowledge
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- 60.0
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Math
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- 65.1
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Mythos Preview
- Not measured
- Claude Opus 4.7 (Adaptive)
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Claude Mythos Preview | Claude Opus 4.7 (Adaptive) | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | Not measured | 75.1 | Not comparable0 vs 3 rows | Not comparable |
| Coding | Not measured | 78.6 | Not comparable0 vs 2 rows | Not comparable |
| Reasoning | Not measured | 75.8 | Not comparable0 vs 1 rows | Not comparable |
| Knowledge | Not measured | 60.0 | Not comparable0 vs 2 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 | 65.1 | Not comparable0 vs 2 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.
A shared-evidence shape is not available.
BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.
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 Mythos Preview
- $0.0875
- Fit state unavailable
- Claude Opus 4.7 (Adaptive)
- $0.0175
- Fits in one request
Claude Opus 4.7 (Adaptive) has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Claude Mythos Preview
- $1.63
- Fit state unavailable
- Claude Opus 4.7 (Adaptive)
- $0.325
- Fits in one request
Claude Opus 4.7 (Adaptive) 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 Mythos Preview
- $6.75
- Fit state unavailable
- Cached input priced at the published list-input rate
- Claude Opus 4.7 (Adaptive)
- $1.35
- Fits in one request
- Cached input priced at the published list-input rate
Claude Opus 4.7 (Adaptive) has the lower modeled cost
Claude Mythos Preview has no published cached-input rate, so cached tokens use its listed input rate. Claude Opus 4.7 (Adaptive) 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 Mythos PreviewClaude Opus 4.7 (Adaptive)
Context window
Maximum documented context; output-token limits may be lower.
Claude Mythos Preview
Not sourced
Claude Opus 4.7 (Adaptive)
1M
API model ID
Claude Mythos Preview
claude-mythos-preview
Anthropic Project GlasswingClaude Opus 4.7 (Adaptive)
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Mythos Preview
Not published
Anthropic Project GlasswingClaude Opus 4.7 (Adaptive)
Not published
Documented inputs
Claude Mythos Preview
Not sourced
Claude Opus 4.7 (Adaptive)
Not sourced
Documented outputs
Claude Mythos Preview
Not sourced
Claude Opus 4.7 (Adaptive)
Not sourced
Provider availability
Claude Mythos Preview
Not sourced
Claude Opus 4.7 (Adaptive)
Not sourced
Reasoning profile
Claude Mythos Preview
Reasoning
Claude Opus 4.7 (Adaptive)
Reasoning
Weight access
Claude Mythos Preview
Proprietary
Claude Opus 4.7 (Adaptive)
Proprietary
License
Claude Mythos Preview
Proprietary
Claude Opus 4.7 (Adaptive)
Proprietary
Release date
Claude Mythos Preview
2026-04-07
Claude Opus 4.7 (Adaptive)
2026-04-16
If you already use one of these models
- Deployment change
- Both entries list Anthropic as the provider. Confirm endpoint, model ID, limits, and feature support before switching.
- 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
- Repository review: $1.63 vs $0.325. Cache-heavy agent loop: $6.75 vs $1.35.
- Context tradeoff
- A complete documented context comparison is not available.
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 evidence22 rows
Agentic
CyberGym
Claude Mythos Preview83.1%SourceClaude Opus 4.7 (Adaptive)73.1%SourceClaude Mythos Preview leads this result
ExploitGym
Claude Mythos Preview17.5%SourceClaude Opus 4.7 (Adaptive)—Not directly comparable
Terminal-Bench 2.0
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)69.4%SourceNot directly comparable
BrowseComp
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)79.3%SourceNot directly comparable
MCP Atlas
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)77.3%SourceNot directly comparable
OSWorld-Verified
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)78%SourceNot directly comparable
OSWorld 2.0
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)18.2%SourceNot directly comparable
JobBench
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)45.9%SourceNot directly comparable
Coding
Reasoning
Knowledge
GPQA
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)94.2%SourceNot directly comparable
GPQA-D
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)94.2%SourceNot directly comparable
HLE
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)54.7%SourceNot directly comparable
HLE w/o tools
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)46.9%SourceNot directly comparable
Math
FrontierMath (legacy)
Claude Mythos Preview—Claude Opus 4.7 (Adaptive)43.8%SourceNot directly comparable
Multimodal
Frequently asked questions
Which is better, Claude Mythos Preview or Claude Opus 4.7 (Adaptive)?
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, Claude Mythos Preview or Claude Opus 4.7 (Adaptive)?
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 Mythos Preview or Claude Opus 4.7 (Adaptive)?
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 Mythos Preview or Claude Opus 4.7 (Adaptive)?
For the stated presets, chat costs $0.0875 on Claude Mythos Preview and $0.0175 on Claude Opus 4.7 (Adaptive); repository review costs $1.63 and $0.325; the cache-heavy agent loop costs $6.75 and $1.35. Claude Mythos Preview has no published cached-input rate, so cached tokens use its listed input rate. Claude Opus 4.7 (Adaptive) has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Claude Mythos Preview or Claude Opus 4.7 (Adaptive)?
A complete documented context-window comparison is not available.
Related comparisons
Last updated August 13, 2026
Watch Claude Mythos Preview vs Claude Opus 4.7 (Adaptive)
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.