A
Anthropic
Claude Mythos 5
- Model type
- Proprietary
- Evidence state
- Supported evidence
- Published context
- 1,000,000
Claude Mythos 5 vs
GPT-5.4
Read the published evidence, route context, and missing facts before making a local decision. This page does not name a universal winner.
Models in this comparison
Provider identity and evidence state stay balanced across the pair.
O
OpenAI
GPT-5.4
- Model type
- Proprietary
- Evidence state
- Supported evidence
- Published context
- 1,050,000
Key implications
Limited shared-metric coverage. Each finding is tied to a published metric, price, context, or modality fact.
- On Agentic, Claude Mythos 5 has a higher supported BenchLM score (75.79 vs 57.52).
- On Overall, Claude Mythos 5 has a higher supported BenchLM score (83.4 vs 73.56).
- Input API price: GPT-5.4 has the lower verified rate ($2.5 / 1M tokens vs $10 / 1M tokens).
- Output API price: GPT-5.4 has the lower verified rate ($15 / 1M tokens vs $50 / 1M tokens).
Shared metric view
A radar is shown only when at least four compatible supported score metrics are published.
Comparable metric detail
- AgenticClaude Mythos 5: 75.79 · GPT-5.4: 57.52
- CodingClaude Mythos 5: 81.66 · GPT-5.4: 58.9
- KnowledgeClaude Mythos 5: 97.4 · GPT-5.4: 75.3
- MathClaude Mythos 5: Unavailable · GPT-5.4: 65.7
- MultimodalGroundedClaude Mythos 5: 84.8 · GPT-5.4: 63.5
- ReasoningClaude Mythos 5: Unavailable · GPT-5.4: 69.8
- OverallClaude Mythos 5: 83.4 · GPT-5.4: 73.56
Source metrics
Friendly metric names and published units stay visible. Missing measurements remain unavailable rather than becoming a score.
| Metric | Unit | Claude Mythos 5 | GPT-5.4 |
|---|---|---|---|
| Agentic | score | 75.79 | 57.52 |
| Coding | score | 81.66 | 58.9 |
| Knowledge | score | 97.4 | 75.3 |
| Math | score | Unavailable | 65.7 |
| MultimodalGrounded | score | 84.8 | 63.5 |
| Reasoning | score | Unavailable | 69.8 |
| Overall | score | 83.4 | 73.56 |
Agentic
- Unit
- score
- Claude Mythos 5
- 75.79
- GPT-5.4
- 57.52
Coding
- Unit
- score
- Claude Mythos 5
- 81.66
- GPT-5.4
- 58.9
Knowledge
- Unit
- score
- Claude Mythos 5
- 97.4
- GPT-5.4
- 75.3
Math
- Unit
- score
- Claude Mythos 5
- Unavailable
- GPT-5.4
- 65.7
MultimodalGrounded
- Unit
- score
- Claude Mythos 5
- 84.8
- GPT-5.4
- 63.5
Reasoning
- Unit
- score
- Claude Mythos 5
- Unavailable
- GPT-5.4
- 69.8
Overall
- Unit
- score
- Claude Mythos 5
- 83.4
- GPT-5.4
- 73.56
Pricing and context
Verification is shown beside each selected route. Missing facts remain Not verified.
| Field | Unit | Claude Mythos 5 | GPT-5.4 |
|---|---|---|---|
| Input API price | USD / 1M tokens | $10 | $2.5 |
| Cached input API price | USD / 1M tokens | $1 | $0.25 |
| Output API price | USD / 1M tokens | $50 | $15 |
| Route context | tokens | 1,000,000 | 1,050,000 |
| Input modalities | published list | Not verified | Not verified |
| Output modalities | published list | Not verified | Not verified |
Input API price
- Unit
- USD / 1M tokens
- Claude Mythos 5
- $10
- GPT-5.4
- $2.5
Cached input API price
- Unit
- USD / 1M tokens
- Claude Mythos 5
- $1
- GPT-5.4
- $0.25
Output API price
- Unit
- USD / 1M tokens
- Claude Mythos 5
- $50
- GPT-5.4
- $15
Route context
- Unit
- tokens
- Claude Mythos 5
- 1,000,000
- GPT-5.4
- 1,050,000
Input modalities
- Unit
- published list
- Claude Mythos 5
- Not verified
- GPT-5.4
- Not verified
Output modalities
- Unit
- published list
- Claude Mythos 5
- Not verified
- GPT-5.4
- Not verified
Evidence provenance
Source records, route identity, timestamps, and methodology are consolidated here without declaring either model a winner.
- Publication time
- Aug 30, 2026, 2:15 AM UTC
- Freshness
- Stale — Published weekly benchmark evidence has not refreshed within 8 days.
- Methodology
- benchlm: benchlm_raw_composite
- Model records
- Claude Mythos 5 — source benchlm · artifact models · model claude-mythos-5
- GPT-5.4 — source benchlm · artifact models · model gpt-5-4
- Selected price routes
- Claude Mythos 5 — route benchlm:claude-mythos-5 · source benchlm · provider anthropic
- GPT-5.4 — route benchlm:gpt-5-4 · source benchlm · provider openai
Switch model pair
Choose from this result’s current and reviewed related models. Switching opens a reviewed comparison; it does not change this result’s evidence.
Step 1
Start with popular models, or search the full selectable directory.
Step 2
Start with popular models, or search the full selectable directory.