Knowledge
Like-for-like- Kimi K2.6
- 42.2
- Ornith-1.5-397B
- 51.1
- Weighted basis
- 2 vs 2 rows
- Reading
- Ornith-1.5-397B leads
Kimi K2.6 vs Ornith-1.5-397B
Updated August 19, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Ornith-1.5-397B has the higher public score estimate, 68.5 versus 60.28, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
10 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
Ornith-1.5-397B
Ornith-1.5-397B 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
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
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: rate-fallback
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.
22 Kimi K2.6 only
10 Shared
8 Ornith-1.5-397B only
- Shared results
- 10
- Kimi K2.6 only
- 22
- Ornith-1.5-397B only
- 8
- Like-for-like categories
- 1 / 8
2 categories use 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
Directional only- Kimi K2.6
- 73.5
- Ornith-1.5-397B
- 86.6
- Weighted basis
- 3 vs 1 rows
- Reading
- Directional only
Coding
Directional only- Kimi K2.6
- 64.4
- Ornith-1.5-397B
- 78.0
- Weighted basis
- 3 vs 2 rows
- Reading
- Directional only
Reasoning
Not comparable- Kimi K2.6
- Not measured
- Ornith-1.5-397B
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- Kimi K2.6
- 67.1
- Ornith-1.5-397B
- Not measured
- Weighted basis
- 4 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Kimi K2.6
- Not measured
- Ornith-1.5-397B
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Kimi K2.6
- 79.8
- Ornith-1.5-397B
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Instruction following
Not comparable- Kimi K2.6
- Not measured
- Ornith-1.5-397B
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Kimi K2.6 | Ornith-1.5-397B | Weighted basis | Reading |
|---|---|---|---|---|
| Knowledge | 42.2 | 51.1 | Like-for-like2 vs 2 rows | Ornith-1.5-397B leads |
| Agentic | 73.5 | 86.6 | Directional only3 vs 1 rows | Directional only |
| Coding | 64.4 | 78.0 | Directional only3 vs 2 rows | Directional only |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Math | 67.1 | Not measured | Not comparable4 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | 79.8 | Not measured | Not comparable2 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.
HLE
Knowledge
SWE-bench Pro
Coding
SWE-bench Verified
Coding
BrowseComp
Agentic
GPQA
Knowledge
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
- Kimi K2.6
- $0.00295
- Fits in one request
- Ornith-1.5-397B
- Self-hosted; infrastructure cost varies
- Fits in one request
Ornith-1.5-397B has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Kimi K2.6
- $0.0595
- Fits in one request
- Ornith-1.5-397B
- Self-hosted; infrastructure cost varies
- Fits in one request
Ornith-1.5-397B has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Kimi K2.6
- $0.249
- Fits in one request
- Cached input priced at the published list-input rate
- Ornith-1.5-397B
- Self-hosted; infrastructure cost varies
- Fits in one request
- Cached-input rate unavailable
Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate. Ornith-1.5-397B 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.
SpecificationKimi K2.6Ornith-1.5-397B
Context window
Maximum documented context; output-token limits may be lower.
Kimi K2.6
256K
Ornith-1.5-397B
API model ID
Kimi K2.6
Not sourced
Ornith-1.5-397B
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Kimi K2.6
Not published
Ornith-1.5-397B
No comparable hosted API rate
Ornith-1.5-397B model cardDocumented inputs
Kimi K2.6
Not sourced
Ornith-1.5-397B
Not sourced
Documented outputs
Kimi K2.6
Not sourced
Ornith-1.5-397B
Not sourced
Provider availability
Kimi K2.6
Not sourced
Ornith-1.5-397B
Not sourced
Reasoning profile
Kimi K2.6
Reasoning
Ornith-1.5-397B
Reasoning
Weight access
Kimi K2.6
Open Weight
Ornith-1.5-397B
Open Weight
License
Kimi K2.6
Open Weight
Ornith-1.5-397B
Open Weight
Release date
Kimi K2.6
2026-04-20
Ornith-1.5-397B
2026-08-18
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
- Ornith-1.5-397B has the higher public score estimate, 68.5 versus 60.28, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Ornith-1.5-397B has the larger documented window (262K).
Run the same representative tasks against both endpoints before changing production traffic.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/day
Ornith-1.5-397B
API / mo$0
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence40 rows
Agentic
Terminal-Bench 2.0
Kimi K2.666.7%SourceOrnith-1.5-397B—Not directly comparable
BrowseComp
Kimi K2.683.2%SourceOrnith-1.5-397B86.6%SourceOrnith-1.5-397B leads this result
OSWorld-Verified
Kimi K2.673.1%SourceOrnith-1.5-397B—Not directly comparable
Toolathlon
Kimi K2.650%SourceOrnith-1.5-397B—Not directly comparable
MCP Atlas
Kimi K2.655.9%SourceOrnith-1.5-397B80%SourceOrnith-1.5-397B leads this result
Claw-Eval
Kimi K2.662.3%SourceOrnith-1.5-397B81.4%SourceOrnith-1.5-397B leads this result
DeepSearchQA
Kimi K2.692.5%SourceOrnith-1.5-397B—Not directly comparable
WideResearch
Kimi K2.680.8%SourceOrnith-1.5-397B80.8%SourceTie
Gert Labs
Kimi K2.656.82%SourceOrnith-1.5-397B—Not directly comparable
ResearchClawBench
Kimi K2.618.0%SourceOrnith-1.5-397B—Not directly comparable
OSWorld 2.0
Kimi K2.64.6%SourceOrnith-1.5-397B—Not directly comparable
Terminal-Bench 2.1
Kimi K2.6—Ornith-1.5-397B86.1%SourceNot directly comparable
HLE w/ tools
Kimi K2.6—Ornith-1.5-397B56.1%SourceNot directly comparable
Toolathlon-Verified
Kimi K2.6—Ornith-1.5-397B71.2%SourceNot directly comparable
Coding
SWE-bench Verified
Kimi K2.680.2%SourceOrnith-1.5-397B86%SourceOrnith-1.5-397B leads this result
LiveCodeBench v6
Kimi K2.689.6%SourceOrnith-1.5-397B—Not directly comparable
SWE-bench Pro
Kimi K2.658.6%SourceOrnith-1.5-397B65.1%SourceOrnith-1.5-397B leads this result
SWE Multilingual
Kimi K2.676.7%SourceOrnith-1.5-397B79.6%SourceOrnith-1.5-397B leads this result
SciCode
Kimi K2.652.2%SourceOrnith-1.5-397B—Not directly comparable
Terminal-Bench 2.0
Kimi K2.666.7%SourceOrnith-1.5-397B—Not directly comparable
Vibe Code Bench
Kimi K2.637.89%SourceOrnith-1.5-397B—Not directly comparable
cursorBench31
Kimi K2.647.6%SourceOrnith-1.5-397B—Not directly comparable
Terminal-Bench 2.1
Kimi K2.6—Ornith-1.5-397B86.1%SourceNot directly comparable
deepSwe
Kimi K2.6—Ornith-1.5-397B56%SourceNot directly comparable
frontierBench
Kimi K2.6—Ornith-1.5-397B13.5%SourceNot directly comparable
NL2Repo
Kimi K2.6—Ornith-1.5-397B59.5%SourceNot directly comparable
Knowledge
GPQA
Kimi K2.690.5%SourceOrnith-1.5-397B92.8%SourceOrnith-1.5-397B leads this result
GPQA-D
Kimi K2.690.5%SourceOrnith-1.5-397B92.8%SourceOrnith-1.5-397B leads this result
HLE
Kimi K2.634.7%SourceOrnith-1.5-397B44.6%SourceOrnith-1.5-397B leads this result
HLE w/o tools
Kimi K2.6—Ornith-1.5-397B44.6%SourceNot directly comparable
Math
AIME26
Kimi K2.696.4%SourceOrnith-1.5-397B—Not directly comparable
HMMT Feb 2026
Kimi K2.692.7%SourceOrnith-1.5-397B—Not directly comparable
MMAnswerBench
Kimi K2.686.0%SourceOrnith-1.5-397B—Not directly comparable
FrontierMath v2 (Tiers 1-3)
Kimi K2.638.966%SourceOrnith-1.5-397B—Not directly comparable
FrontierMath v2 (Tier 4)
Kimi K2.614.580%SourceOrnith-1.5-397B—Not directly comparable
Multimodal
MMMU-Pro
Kimi K2.679.4%SourceOrnith-1.5-397B—Not directly comparable
MMMU-Pro w/ Python
Kimi K2.680.1%SourceOrnith-1.5-397B—Not directly comparable
CharXiv
Kimi K2.680.4%SourceOrnith-1.5-397B—Not directly comparable
MathVision
Kimi K2.687.4%SourceOrnith-1.5-397B—Not directly comparable
V*
Kimi K2.696.9%SourceOrnith-1.5-397B—Not directly comparable
Frequently asked questions
Which is better, Kimi K2.6 or Ornith-1.5-397B?
Ornith-1.5-397B has the higher public score estimate, 68.5 versus 60.28, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Kimi K2.6 or Ornith-1.5-397B?
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, Kimi K2.6 or Ornith-1.5-397B?
The current agentic tasks 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 costs less, Kimi K2.6 or Ornith-1.5-397B?
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, Kimi K2.6 or Ornith-1.5-397B?
Ornith-1.5-397B has the larger documented context window: 262K, compared with 256K.
Related comparisons
Last updated August 19, 2026
Watch Kimi K2.6 vs Ornith-1.5-397B
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.