Coding
Like-for-like- Granite 4.2 8B
- 36.5
- Kimi K2.6
- 64.4
- Weighted basis
- 3 vs 3 rows
- Reading
- Kimi K2.6 leads
Granite 4.2 8B vs Kimi K2.6
Updated August 31, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Kimi K2.6 has the higher public score estimate, 60.06 versus 46.88, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
6 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.
Coding work
Code generation, repair, and software-engineering tasks
Kimi K2.6
Kimi K2.6 leads on the same 3 weighted benchmark rows.
Confidence: stronger
Long documents
Prompts that approach the documented context limit
Kimi K2.6
Kimi K2.6 has the larger documented context window.
Confidence: documented
Show secondary and unsupported calls
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
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Granite 4.2 8B does not fit this workload in one request. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate. Granite 4.2 8B has no comparable published API token rate.
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.
8 Granite 4.2 8B only
6 Shared
26 Kimi K2.6 only
- Shared results
- 6
- Granite 4.2 8B only
- 8
- Kimi K2.6 only
- 26
- Like-for-like categories
- 1 / 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.
Knowledge
Directional only- Granite 4.2 8B
- 72.2
- Kimi K2.6
- 42.2
- Weighted basis
- 2 vs 2 rows
- Reading
- Directional only
Agentic
Not comparable- Granite 4.2 8B
- Not measured
- Kimi K2.6
- 73.5
- Weighted basis
- 0 vs 3 rows
- Reading
- Not comparable
Reasoning
Not comparable- Granite 4.2 8B
- Not measured
- Kimi K2.6
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- Granite 4.2 8B
- Not measured
- Kimi K2.6
- 67.1
- Weighted basis
- 0 vs 4 rows
- Reading
- Not comparable
Multilingual
Not comparable- Granite 4.2 8B
- Not measured
- Kimi K2.6
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Granite 4.2 8B
- Not measured
- Kimi K2.6
- 79.8
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- Granite 4.2 8B
- 79.3
- Kimi K2.6
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
| Category | Granite 4.2 8B | Kimi K2.6 | Weighted basis | Reading |
|---|---|---|---|---|
| Coding | 36.5 | 64.4 | Like-for-like3 vs 3 rows | Kimi K2.6 leads |
| Knowledge | 72.2 | 42.2 | Directional only2 vs 2 rows | Directional only |
| Agentic | Not measured | 73.5 | Not comparable0 vs 3 rows | Not comparable |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Math | Not measured | 67.1 | Not comparable0 vs 4 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 79.8 | Not comparable0 vs 2 rows | Not comparable |
| Instruction following | 79.3 | Not measured | Not comparable1 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 Pro
Coding
SWE-bench Verified
Coding
GPQA
Knowledge
SciCode
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
- Granite 4.2 8B
- Self-hosted; infrastructure cost varies
- Fits in one request
- Kimi K2.6
- $0.00295
- Fits in one request
Granite 4.2 8B has no comparable published API token rate.
Repository review
50K fresh input + 3K output tokens
- Granite 4.2 8B
- Self-hosted; infrastructure cost varies
- Fits in one request
- Kimi K2.6
- $0.0595
- Fits in one request
Granite 4.2 8B has no comparable published API token rate.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Granite 4.2 8B
- Self-hosted; infrastructure cost varies
- Does not fit in one request
- Cached-input rate unavailable
- Kimi K2.6
- $0.249
- Fits in one request
- Cached input priced at the published list-input rate
Granite 4.2 8B does not fit this workload in one request. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate. Granite 4.2 8B 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.
SpecificationGranite 4.2 8BKimi K2.6
Context window
Maximum documented context; output-token limits may be lower.
Granite 4.2 8B
Kimi K2.6
256K
API model ID
Granite 4.2 8B
ibm-granite/granite-4.2-8b
IBM Granite 4.2 8B model cardKimi K2.6
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Granite 4.2 8B
No comparable hosted API rate
IBM Granite 4.2 8B model cardKimi K2.6
Not published
Documented inputs
Granite 4.2 8B
Not sourced
Kimi K2.6
Not sourced
Documented outputs
Granite 4.2 8B
Not sourced
Kimi K2.6
Not sourced
Provider availability
Granite 4.2 8B
Not sourced
Kimi K2.6
Not sourced
Reasoning profile
Granite 4.2 8B
Reasoning
Kimi K2.6
Reasoning
Weight access
Granite 4.2 8B
Open Weight
Kimi K2.6
Open Weight
License
Granite 4.2 8B
Open Weight
Kimi K2.6
Open Weight
Release date
Granite 4.2 8B
2026-08-25
Kimi K2.6
2026-04-20
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
- Kimi K2.6 has the higher public score estimate, 60.06 versus 46.88, but the 90% score intervals overlap.
- Workload cost
- A complete comparable API-rate estimate is not available for both models.
- Context tradeoff
- Kimi K2.6 has the larger documented window (256K).
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.
Granite 4.2 8B
API / mo$0
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/day
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 2.1
Granite 4.2 8B20.6%SourceKimi K2.6—Not directly comparable
τ³-bench results
Granite 4.2 8B58.1%SourceKimi K2.6—Not directly comparable
BFCL v4
Granite 4.2 8B52.4%SourceKimi K2.6—Not directly comparable
Terminal-Bench 2.0
Granite 4.2 8B—Kimi K2.666.7%SourceNot directly comparable
BrowseComp
Granite 4.2 8B—Kimi K2.683.2%SourceNot directly comparable
OSWorld-Verified
Granite 4.2 8B—Kimi K2.673.1%SourceNot directly comparable
Toolathlon
Granite 4.2 8B—Kimi K2.650%SourceNot directly comparable
MCP Atlas
Granite 4.2 8B—Kimi K2.655.9%SourceNot directly comparable
Claw-Eval
Granite 4.2 8B—Kimi K2.662.3%SourceNot directly comparable
DeepSearchQA
Granite 4.2 8B—Kimi K2.692.5%SourceNot directly comparable
WideResearch
Granite 4.2 8B—Kimi K2.680.8%SourceNot directly comparable
Gert Labs
Granite 4.2 8B—Kimi K2.656.82%SourceNot directly comparable
ResearchClawBench
Granite 4.2 8B—Kimi K2.618.0%SourceNot directly comparable
OSWorld 2.0
Granite 4.2 8B—Kimi K2.64.6%SourceNot directly comparable
Coding
SWE-bench Verified
Granite 4.2 8B47.7%SourceKimi K2.680.2%SourceKimi K2.6 leads this result
SWE-bench Pro
Granite 4.2 8B19.1%SourceKimi K2.658.6%SourceKimi K2.6 leads this result
SWE Multilingual
Granite 4.2 8B30.8%SourceKimi K2.676.7%SourceKimi K2.6 leads this result
Terminal-Bench 2.1
Granite 4.2 8B20.6%SourceKimi K2.6—Not directly comparable
LiveCodeBench v6
Granite 4.2 8B73.2%SourceKimi K2.689.6%SourceKimi K2.6 leads this result
SciCode
Granite 4.2 8B36.1%SourceKimi K2.652.2%SourceKimi K2.6 leads this result
Terminal-Bench 2.0
Granite 4.2 8B—Kimi K2.666.7%SourceNot directly comparable
Vibe Code Bench
Granite 4.2 8B—Kimi K2.637.89%SourceNot directly comparable
cursorBench31
Granite 4.2 8B—Kimi K2.647.6%SourceNot directly comparable
Knowledge
Math
AIME 2025
Granite 4.2 8B86.7%SourceKimi K2.6—Not directly comparable
HMMT Feb 2025
Granite 4.2 8B78.3%SourceKimi K2.6—Not directly comparable
AIME26
Granite 4.2 8B—Kimi K2.696.4%SourceNot directly comparable
HMMT Feb 2026
Granite 4.2 8B—Kimi K2.692.7%SourceNot directly comparable
MMAnswerBench
Granite 4.2 8B—Kimi K2.686.0%SourceNot directly comparable
FrontierMath v2 (Tiers 1-3)
Granite 4.2 8B—Kimi K2.638.966%SourceNot directly comparable
FrontierMath v2 (Tier 4)
Granite 4.2 8B—Kimi K2.614.580%SourceNot directly comparable
Multimodal
MMMU-Pro
Granite 4.2 8B—Kimi K2.679.4%SourceNot directly comparable
MMMU-Pro w/ Python
Granite 4.2 8B—Kimi K2.680.1%SourceNot directly comparable
CharXiv
Granite 4.2 8B—Kimi K2.680.4%SourceNot directly comparable
MathVision
Granite 4.2 8B—Kimi K2.687.4%SourceNot directly comparable
V*
Granite 4.2 8B—Kimi K2.696.9%SourceNot directly comparable
Instruction following
IFBench
Granite 4.2 8B79.3%SourceKimi K2.6—Not directly comparable
Frequently asked questions
Which is better, Granite 4.2 8B or Kimi K2.6?
Kimi K2.6 has the higher public score estimate, 60.06 versus 46.88, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Granite 4.2 8B or Kimi K2.6?
Kimi K2.6 leads the like-for-like coding comparison across 3 shared weighted benchmark rows.
Which is better for agentic tasks, Granite 4.2 8B or Kimi K2.6?
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, Granite 4.2 8B or Kimi K2.6?
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, Granite 4.2 8B or Kimi K2.6?
Kimi K2.6 has the larger documented context window: 256K, compared with 128K.
Related comparisons
Last updated August 31, 2026
Watch Granite 4.2 8B vs Kimi K2.6
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.