Skip to main content
Radar

Five or fewer confirmed AI changes, with original sources, on mornings when something changed.A free source-linked morning brief.

Start free brief
BenchLM
Model A
Kimi K2.5

Moonshot AI

58.9/100

Supported · Public rank #74

90% interval 51.0–66.8

Kimi K2.5 vs Toast 1

Updated August 14, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B
Toast 1

Mixedbread

—

Evidence status unavailable

90% interval unavailable

Decision reading

The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. Use the documented cost, context, and runtime rows instead.

0 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

    Kimi K2.5

    Kimi K2.5 has the larger documented context window.

    Confidence: documented

  • Chat turn cost

    1K fresh input + 500 output tokens

    Toast 1

    Toast 1 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.

    Confidence: listed-rates

Show secondary and unsupported calls

  • Repository review cost

    50K fresh input + 3K output tokens

    Toast 1

    Toast 1 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

  • 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. Toast 1 does not fit this workload in one request. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.

    Confidence: rate-fallback

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

Shared results
0
Kimi K2.5 only
45
Toast 1 only
0
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.

Agentic

Not comparable
Kimi K2.5
55.0
Toast 1
Not measured
Weighted basis
2 vs 0 rows
Reading
Not comparable

Coding

Not comparable
Kimi K2.5
59.4
Toast 1
Not measured
Weighted basis
4 vs 0 rows
Reading
Not comparable

Reasoning

Not comparable
Kimi K2.5
61.0
Toast 1
Not measured
Weighted basis
1 vs 0 rows
Reading
Not comparable

Knowledge

Not comparable
Kimi K2.5
56.9
Toast 1
Not measured
Weighted basis
4 vs 0 rows
Reading
Not comparable

Math

Not comparable
Kimi K2.5
60.6
Toast 1
Not measured
Weighted basis
4 vs 0 rows
Reading
Not comparable

Multilingual

Not comparable
Kimi K2.5
82.3
Toast 1
Not measured
Weighted basis
1 vs 0 rows
Reading
Not comparable

Multimodal

Not comparable
Kimi K2.5
78.5
Toast 1
Not measured
Weighted basis
1 vs 0 rows
Reading
Not comparable

Instruction following

Not comparable
Kimi K2.5
93.9
Toast 1
Not measured
Weighted basis
1 vs 0 rows
Reading
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

Kimi K2.5
$0.0021
Fits in one request
Toast 1
$0.00066
Fits in one request

Toast 1 has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

Kimi K2.5
$0.039
Fits in one request
Toast 1
$0.01716
Fits in one request

Toast 1 has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

Kimi K2.5
$0.162
Fits in one request
Cached input priced at the published list-input rate
Toast 1
$0.0204
Does not fit in one request

Toast 1 does not fit this workload in one request. Kimi K2.5 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.

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

Kimi K2.5

Not published

Toast 1

$0.036 per 1M cached input tokens

Mixedbread pricing

Reasoning profile

Kimi K2.5

Non-Reasoning

Toast 1

Reasoning

Weight access

Kimi K2.5

Open Weight

Toast 1

Proprietary

License

Kimi K2.5

Open Weight

Toast 1

Proprietary

Release date

Kimi K2.5

2026-02-01

Toast 1

2026-08-13

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
The public evidence has no benchmark result shared by both models, so it does not support a quality verdict.
Workload cost
Repository review: $0.039 vs $0.01716. Cache-heavy agent loop: $0.162 vs $0.0204.
Context tradeoff
Kimi K2.5 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.

Kimi K2.5
API / mo$2,700
Self-host / mo$5,221
Break-even132M/day
Toast 1
API / mo$765
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Model the full break-even

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence45 rows

Agentic

  • Terminal-Bench 2.0

    Kimi K2.550.8%
    Source
    Toast 1—

    Not directly comparable

  • BrowseComp

    Kimi K2.560.6%
    Source
    Toast 1—

    Not directly comparable

  • Claw-Eval

    Kimi K2.552.3%
    Source
    Toast 1—

    Not directly comparable

  • QwenClawBench

    Kimi K2.554.3%
    Source
    Toast 1—

    Not directly comparable

  • τ³-bench results

    Kimi K2.565.7%
    Source
    Toast 1—

    Not directly comparable

  • DeepSearchQA

    Kimi K2.577.1%
    Source
    Toast 1—

    Not directly comparable

  • DeepPlanning

    Kimi K2.514.4%
    Source
    Toast 1—

    Not directly comparable

  • Toolathlon

    Kimi K2.527.8%
    Source
    Toast 1—

    Not directly comparable

  • MCP Atlas

    Kimi K2.529.5%
    Source
    Toast 1—

    Not directly comparable

  • MCP-Tasks

    Kimi K2.559.1%
    Source
    Toast 1—

    Not directly comparable

  • WideResearch

    Kimi K2.572.7%
    Source
    Toast 1—

    Not directly comparable

  • Gert Labs

    Kimi K2.545.88%
    Source
    Toast 1—

    Not directly comparable

  • ResearchClawBench

    Kimi K2.514.0%
    Source
    Toast 1—

    Not directly comparable

  • JobBench

    Kimi K2.58.7%
    Source
    Toast 1—

    Not directly comparable

Coding

  • SWE-bench Verified

    Kimi K2.576.8%
    Source
    Toast 1—

    Not directly comparable

  • SWE-bench Verified*

    Kimi K2.570.8%
    Source
    Toast 1—

    Not directly comparable

  • LiveCodeBench v6

    Kimi K2.585.0%
    Source
    Toast 1—

    Not directly comparable

  • SWE-bench Pro

    Kimi K2.550.7%
    Source
    Toast 1—

    Not directly comparable

  • SWE Multilingual

    Kimi K2.573%
    Source
    Toast 1—

    Not directly comparable

  • SWE-Rebench

    Kimi K2.558.5%
    Source
    Toast 1—

    Not directly comparable

  • React Native Evals

    Kimi K2.577.2%
    Source
    Toast 1—

    Not directly comparable

  • SciCode

    Kimi K2.548.7%
    Source
    Toast 1—

    Not directly comparable

Reasoning

  • LongBench v2

    Kimi K2.561%
    Source
    Toast 1—

    Not directly comparable

Knowledge

  • GPQA

    Kimi K2.587.6%
    Source
    Toast 1—

    Not directly comparable

  • GPQA-D

    Kimi K2.587.6%
    Source
    Toast 1—

    Not directly comparable

  • SuperGPQA

    Kimi K2.569.2%
    Source
    Toast 1—

    Not directly comparable

  • MMLU-Pro

    Kimi K2.587.1%
    Source
    Toast 1—

    Not directly comparable

  • MMLU-Pro (Arcee)

    Kimi K2.587.1%
    Source
    Toast 1—

    Not directly comparable

  • HLE

    Kimi K2.530.1%
    Source
    Toast 1—

    Not directly comparable

Math

  • AIME 2025

    Kimi K2.596.1%
    Source
    Toast 1—

    Not directly comparable

  • AIME26

    Kimi K2.595.8%
    Source
    Toast 1—

    Not directly comparable

  • AIME25 (Arcee)

    Kimi K2.596.3%
    Source
    Toast 1—

    Not directly comparable

  • HMMT Feb 2025

    Kimi K2.595.4%
    Source
    Toast 1—

    Not directly comparable

  • HMMT Nov 2025

    Kimi K2.591.1%
    Source
    Toast 1—

    Not directly comparable

  • HMMT Feb 2026

    Kimi K2.587.1%
    Source
    Toast 1—

    Not directly comparable

  • MMAnswerBench

    Kimi K2.581.8%
    Source
    Toast 1—

    Not directly comparable

  • FrontierMath v2 (Tiers 1-3)

    Kimi K2.527.900%
    Source
    Toast 1—

    Not directly comparable

  • FrontierMath v2 (Tier 4)

    Kimi K2.54.200%
    Source
    Toast 1—

    Not directly comparable

Multilingual

  • MMLU-ProX

    Kimi K2.582.3%
    Source
    Toast 1—

    Not directly comparable

  • NOVA-63

    Kimi K2.556.0%
    Source
    Toast 1—

    Not directly comparable

Multimodal

  • MMMU-Pro

    Kimi K2.578.5%
    Source
    Toast 1—

    Not directly comparable

  • Video-MME

    Kimi K2.587.4%
    Source
    Toast 1—

    Not directly comparable

  • MMVU

    Kimi K2.580.4%
    Source
    Toast 1—

    Not directly comparable

  • VideoMMMU

    Kimi K2.586.6%
    Source
    Toast 1—

    Not directly comparable

Instruction following

  • IFEval

    Kimi K2.593.9%
    Source
    Toast 1—

    Not directly comparable

Frequently asked questions

Which is better, Kimi K2.5 or Toast 1?

The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. The page therefore keeps the decision tied to the specific documented workload.

Which is better for coding, Kimi K2.5 or Toast 1?

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, Kimi K2.5 or Toast 1?

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, Kimi K2.5 or Toast 1?

For the stated presets, chat costs $0.0021 on Kimi K2.5 and $0.00066 on Toast 1; repository review costs $0.039 and $0.01716; the cache-heavy agent loop costs $0.162 and $0.0204. Toast 1 does not fit this workload in one request. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, Kimi K2.5 or Toast 1?

Kimi K2.5 has the larger documented context window: 256K, compared with 131K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated August 14, 2026

Watch Kimi K2.5 vs Toast 1

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.