M
Moonshot AI
Kimi K3
- Model type
- Unknown
- Evidence state
- Supported evidence
- Published context
- 1,050,000
Kimi K3 vs
Qwen3.8 Max
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.
A
Alibaba
Qwen3.8 Max
- Model type
- Open Weight
- Evidence state
- Supported evidence
- Published context
- 1,000,000
Key implications
Broad shared-metric coverage. Each finding is tied to a published metric, price, context, or modality fact.
- Across compatible supported BenchLM categories, Kimi K3 has higher scores in Knowledge, Coding, and Overall; Qwen3.8 Max has higher scores in Agentic and MultimodalGrounded.
- Context window: Kimi K3 has the larger published context window (1,050,000 tokens vs 1,000,000 tokens).
Shared metric view
The radar is a per-axis relative view; exact values and units remain available in its adjacent table.
- Kimi K3: solid line
- Qwen3.8 Max: dashed line
| Metric | Kimi K3 | Qwen3.8 Max | Unit |
|---|---|---|---|
| Agentic | 73.64 | 75.17 | score |
| Coding | 78.58 | 65.78 | score |
| Knowledge | 83.8 | 64.7 | score |
| MultimodalGrounded | 86.2 | 87.1 | score |
| Overall | 80.61 | 79.22 | score |
Source metrics
Friendly metric names and published units stay visible. Missing measurements remain unavailable rather than becoming a score.
| Metric | Unit | Kimi K3 | Qwen3.8 Max |
|---|---|---|---|
| Agentic | score | 73.64 | 75.17 |
| Coding | score | 78.58 | 65.78 |
| InstructionFollowing | score | Unavailable | 93.9 |
| Knowledge | score | 83.8 | 64.7 |
| MultimodalGrounded | score | 86.2 | 87.1 |
| Reasoning | score | Unavailable | 95.5 |
| Overall | score | 80.61 | 79.22 |
Agentic
- Unit
- score
- Kimi K3
- 73.64
- Qwen3.8 Max
- 75.17
Coding
- Unit
- score
- Kimi K3
- 78.58
- Qwen3.8 Max
- 65.78
InstructionFollowing
- Unit
- score
- Kimi K3
- Unavailable
- Qwen3.8 Max
- 93.9
Knowledge
- Unit
- score
- Kimi K3
- 83.8
- Qwen3.8 Max
- 64.7
MultimodalGrounded
- Unit
- score
- Kimi K3
- 86.2
- Qwen3.8 Max
- 87.1
Reasoning
- Unit
- score
- Kimi K3
- Unavailable
- Qwen3.8 Max
- 95.5
Overall
- Unit
- score
- Kimi K3
- 80.61
- Qwen3.8 Max
- 79.22
Pricing and context
Verification is shown beside each selected route. Missing facts remain Not verified.
| Field | Unit | Kimi K3 | Qwen3.8 Max |
|---|---|---|---|
| Input API price | USD / 1M tokens | $3 | Not verified |
| Cached input API price | USD / 1M tokens | $0.3 | Not verified |
| Output API price | USD / 1M tokens | $15 | Not verified |
| Route context | tokens | 1,050,000 | Not verified |
| Input modalities | published list | Not verified | Not verified |
| Output modalities | published list | Not verified | Not verified |
Input API price
- Unit
- USD / 1M tokens
- Kimi K3
- $3
- Qwen3.8 Max
- Not verified
Cached input API price
- Unit
- USD / 1M tokens
- Kimi K3
- $0.3
- Qwen3.8 Max
- Not verified
Output API price
- Unit
- USD / 1M tokens
- Kimi K3
- $15
- Qwen3.8 Max
- Not verified
Route context
- Unit
- tokens
- Kimi K3
- 1,050,000
- Qwen3.8 Max
- Not verified
Input modalities
- Unit
- published list
- Kimi K3
- Not verified
- Qwen3.8 Max
- Not verified
Output modalities
- Unit
- published list
- Kimi K3
- Not verified
- Qwen3.8 Max
- 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
- Kimi K3 — source benchlm · artifact models · model kimi-3
- Qwen3.8 Max — source benchlm · artifact models · model qwen3-8-max
- Selected price routes
- Kimi K3 — route benchlm:kimi-3 · source benchlm · provider moonshot-ai
- Qwen3.8 Max — Not published
Keep comparing
Other reviewed matchups from the same published revision, ready to open without changing this result’s evidence.
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.