GPT-5.6 Terra vs Kimi K2.5

Updated September 24, 2026. Rank says GPT-5.6 Terra is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.

Decision reading

GPT-5.6 Terra has the higher public score, 72.58 versus 52.96, and the 90% score intervals do not overlap. 8 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Model A

GPT-5.6 Terra

OpenAI

72.58/100

Supported · Public rank #9

90% interval 68.5–76.7

Model B

Kimi K2.5

Moonshot AI

52.96/100

Supported · Public rank #63

90% interval 44.8–61.1

Shared results: 8
GPT-5.6 Terra only: 24
Kimi K2.5 only: 37
Like-for-like categories: 1 / 8

Supported: GPT-5.6 Terra and Kimi K2.5How the comparison works

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
GPT-5.6 Terra
GPT-5.6 Terra has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback

Show secondary and unsupported calls

Repository review cost
50K fresh input + 3K output tokens
Kimi K2.5
Kimi K2.5 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
Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Kimi K2.5 is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limited

Which one for a specific job

Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.

Pick the job you would use them for

The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.

64.7GPT-5.6 Terra38.9Kimi K2.5

Directional only · BenchAlign v5.7

GPT-5.6 Terra scores higher, but at least one score rests on Estimated evidence or a different benchmark set. Directional only, no winner.

Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.

Open the coding leaderboard Shortlist models for this job

Same basis rules as the category table below

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.

FrontierMath v2 (Tier 4)Math
Normalized gap 64.1
GPT-5.6 Terra:68.300%
Kimi K2.5:4.200%
GPT-5.6 Terra source Kimi K2.5 source
FrontierMath v2 (Tiers 1-3)Math
Normalized gap 57.0
GPT-5.6 Terra:84.900%
Kimi K2.5:27.900%
GPT-5.6 Terra source Kimi K2.5 source
BrowseCompAgentic
Normalized gap 26.9
GPT-5.6 Terra:87.5%
Kimi K2.5:60.6%
GPT-5.6 Terra source Kimi K2.5 source
SWE-bench ProCoding
Normalized gap 12.7
GPT-5.6 Terra:63.4%
Kimi K2.5:50.7%
GPT-5.6 Terra source Kimi K2.5 source
GPQAKnowledge
Normalized gap 5.3
GPT-5.6 Terra:92.9%
Kimi K2.5:87.6%
GPT-5.6 Terra source Kimi K2.5 source

Bars run 0–100 on each benchmark’s normalized display scale

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign v5.7 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Multimodal

Like-for-like

GPT-5.6 Terra: 77.2; #16/50
Kimi K2.5: 65.8; #25/50
Basis: Provisional lane · 1 vs 1 weighted rows
Reading: GPT-5.6 Terra leads

Agentic

Directional only

GPT-5.6 Terra: 59.5; Supported · #18/105
Kimi K2.5: 36.1; Estimated · #52/105
Basis: BenchAlign v5.7 lane · 9 vs 14 public rows
Reading: Directional only

Coding

Directional only

GPT-5.6 Terra: 64.7; Supported · #8/135
Kimi K2.5: 38.9; Estimated · #61/135
Basis: BenchAlign v5.7 lane · 8 vs 8 public rows
Reading: Directional only

Knowledge

Directional only

GPT-5.6 Terra: 71.1; Supported · #9/158
Kimi K2.5: 47.8; Estimated · #65/158
Basis: BenchAlign v5.7 lane · 8 vs 6 public rows
Reading: Directional only

Instruction following

Directional only

GPT-5.6 Terra: 85.7; #35/124
Kimi K2.5: 84.4; #41/124
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Directional only

Reasoning

Not comparable

GPT-5.6 Terra: 65.4; #12/19
Kimi K2.5: 55.1; Unranked · 3 rankable rows
Basis: Provisional lane · 2 vs 1 weighted rows
Reading: Not comparable

Multilingual

Not comparable

GPT-5.6 Terra: Not ranked
Kimi K2.5: 38.2; #8/12
Basis: Provisional lane · 0 vs 1 weighted rows
Reading: Not comparable

Math

Not comparable

GPT-5.6 Terra: 96.8; Unranked · 3 rankable rows
Kimi K2.5: 62.1; #4/7
Basis: Provisional lane · 2 vs 4 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for GPT-5.6 Terra and Kimi K2.5
Category	GPT-5.6 Terra	Kimi K2.5	Basis	Reading
Multimodal	77.2#16/50	65.8#25/50	Like-for-likeProvisional lane · 1 vs 1 weighted rows	GPT-5.6 Terra leads
Agentic	59.5Supported · #18/105	36.1Estimated · #52/105	Directional onlyBenchAlign v5.7 lane · 9 vs 14 public rows	Directional only
Coding	64.7Supported · #8/135	38.9Estimated · #61/135	Directional onlyBenchAlign v5.7 lane · 8 vs 8 public rows	Directional only
Knowledge	71.1Supported · #9/158	47.8Estimated · #65/158	Directional onlyBenchAlign v5.7 lane · 8 vs 6 public rows	Directional only
Instruction following	85.7#35/124	84.4#41/124	Directional onlyProvisional lane · 0 vs 0 weighted rows	Directional only
Reasoning	65.4#12/19	55.1Unranked · 3 rankable rows	Not comparableProvisional lane · 2 vs 1 weighted rows	Not comparable
Multilingual	Not ranked	38.2#8/12	Not comparableProvisional lane · 0 vs 1 weighted rows	Not comparable
Math	96.8Unranked · 3 rankable rows	62.1#4/7	Not comparableProvisional lane · 2 vs 4 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

Supported evidence per lane · bars run 0–100Methodology

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

GPT-5.6 Terra: $0.008; Fits in one request
Kimi K2.5: $0.0021; Fits in one request

Kimi K2.5 has the lower modeled cost

Costs use the listed standard API rates.

Repository review

50K fresh input + 3K output tokens

GPT-5.6 Terra: $0.136; Fits in one request
Kimi K2.5: $0.039; Fits in one request

Kimi K2.5 has the lower modeled cost

Costs use the listed standard API rates.

Cache-heavy agent loop

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

GPT-5.6 Terra: $0.2; Fits in one request
Kimi K2.5: $0.162; Fits in one request; Cached input priced at the published list-input rate

Kimi K2.5 has the lower modeled cost

Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.

Cached input falls back to the list input rate only where a cached rate is unpublished

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationGPT-5.6 TerraKimi K2.5

Context window

Maximum documented context; output-token limits may be lower.

GPT-5.6 Terra

1.05M

OpenAI model catalog

Kimi K2.5

256K

API model ID

GPT-5.6 Terra

gpt-5.6-terra

OpenAI model catalog

Kimi K2.5

Not sourced

Cached-input rate

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

GPT-5.6 Terra

$0.2 per 1M cached input tokens

OpenAI pricing

Kimi K2.5

Not published

Documented inputs

GPT-5.6 Terra

text, image

OpenAI model catalog

Kimi K2.5

Not sourced

Documented outputs

GPT-5.6 Terra

text

OpenAI model catalog

Kimi K2.5

Not sourced

Provider availability

GPT-5.6 Terra

Generally Available · OpenAI Responses API

OpenAI model catalog

Kimi K2.5

Not sourced

Reasoning profile

GPT-5.6 Terra

Reasoning

Kimi K2.5

Non-Reasoning

Weight access

GPT-5.6 Terra

Proprietary

Kimi K2.5

Open Weight

License

GPT-5.6 Terra

Proprietary

Kimi K2.5

Open Weight

Release date

GPT-5.6 Terra

2026-07-09

Kimi K2.5

2026-02-01

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: GPT-5.6 Terra has the higher public score, 72.58 versus 52.96, and the 90% score intervals do not overlap.
Workload cost: Repository review: $0.136 vs $0.039. Cache-heavy agent loop: $0.2 vs $0.162.
Context tradeoff: GPT-5.6 Terra has the larger documented window (1.05M).

Run the same representative tasks against both endpoints before changing production traffic.

Questions

Which is better, GPT-5.6 Terra or Kimi K2.5?

GPT-5.6 Terra has the higher public score, 72.58 versus 52.96, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.

Which is better for coding, GPT-5.6 Terra or Kimi K2.5?

GPT-5.6 Terra scores higher for coding on the public lane, 64.7 to 38.9. Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, GPT-5.6 Terra or Kimi K2.5?

GPT-5.6 Terra scores higher for agentic tasks on the public lane, 59.5 to 36.1. Kimi K2.5 is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which costs less, GPT-5.6 Terra or Kimi K2.5?

For the stated presets, chat costs $0.008 on GPT-5.6 Terra and $0.0021 on Kimi K2.5; repository review costs $0.136 and $0.039; the cache-heavy agent loop costs $0.2 and $0.162. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.

Which has the larger context window, GPT-5.6 Terra or Kimi K2.5?

GPT-5.6 Terra has the larger documented context window: 1.05M, compared with 256K.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

GPT-5.6 Terra

API / mo$10,500

Self-host / moNot listed

Break-even—

Proprietary model — self-hosting not applicable.

Kimi K2.5

API / mo$2,700

Self-host / mo$5,221

Break-even132M/day

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 evidence69 rows

Agentic

Terminal-Bench 3.0
GPT-5.6 Terra20.8%
Source
Kimi K2.5—
Not directly comparable
Terminal-Bench 2.1
GPT-5.6 Terra87.4%
Source
Kimi K2.5—
Not directly comparable
BrowseComp
GPT-5.6 Terra87.5%
Source
Kimi K2.560.6%
Source
GPT-5.6 Terra leads this result
OSWorld 2.0
GPT-5.6 Terra50.2%
Source
Kimi K2.5—
Not directly comparable
CyberGym
GPT-5.6 Terra81.8%
Source
Kimi K2.5—
Not directly comparable
ExploitGym
GPT-5.6 Terra23.2%
Source
Kimi K2.5—
Not directly comparable
Toolathlon
GPT-5.6 Terra53.1%
Source
Kimi K2.527.8%
Source
GPT-5.6 Terra leads this result
Terminal-Bench 2.1 (Vals)
GPT-5.6 Terra77.5%
Source
Kimi K2.5—
Not directly comparable
ApprenticeBench
GPT-5.6 Terra16%
Source
Kimi K2.5—
Not directly comparable
Terminal-Bench 2.0
GPT-5.6 Terra—
Kimi K2.550.8%
Source
Not directly comparable
Claw-Eval
GPT-5.6 Terra—
Kimi K2.552.3%
Source
Not directly comparable
QwenClawBench
GPT-5.6 Terra—
Kimi K2.554.3%
Source
Not directly comparable
τ³-bench results
GPT-5.6 Terra—
Kimi K2.565.7%
Source
Not directly comparable
DeepSearchQA
GPT-5.6 Terra—
Kimi K2.577.1%
Source
Not directly comparable
DeepPlanning
GPT-5.6 Terra—
Kimi K2.514.4%
Source
Not directly comparable
MCP Atlas
GPT-5.6 Terra—
Kimi K2.529.5%
Source
Not directly comparable
MCP-Tasks
GPT-5.6 Terra—
Kimi K2.559.1%
Source
Not directly comparable
WideResearch
GPT-5.6 Terra—
Kimi K2.572.7%
Source
Not directly comparable
Gert Labs
GPT-5.6 Terra—
Kimi K2.545.88%
Source
Not directly comparable
ResearchClawBench
GPT-5.6 Terra—
Kimi K2.514.0%
Source
Not directly comparable
JobBench
GPT-5.6 Terra—
Kimi K2.58.7%
Source
Not directly comparable

Coding

SWE-bench Pro
GPT-5.6 Terra63.4%
Source
Kimi K2.550.7%
Source
GPT-5.6 Terra leads this result
Terminal-Bench 2.1
GPT-5.6 Terra87.4%
Source
Kimi K2.5—
Not directly comparable
DeepSWE
GPT-5.6 Terra69.6%
Source
Kimi K2.5—
Not directly comparable
FrontierCode 1.1 Extended
GPT-5.6 Terra55.8%
Source
Kimi K2.5—
Not directly comparable
cursorBench32
GPT-5.6 Terra64.9%
Source
Kimi K2.5—
Not directly comparable
VulcanBench v3
GPT-5.6 Terra87.0%
Source
Kimi K2.5—
Not directly comparable
LiveCodeBench (Vals)
GPT-5.6 Terra85.9%
Source
Kimi K2.5—
Not directly comparable
SWE-bench (Vals)
GPT-5.6 Terra95.4%
Source
Kimi K2.5—
Not directly comparable
SWE-bench Verified
GPT-5.6 Terra—
Kimi K2.576.8%
Source
Not directly comparable
SWE-bench Verified*
GPT-5.6 Terra—
Kimi K2.570.8%
Source
Not directly comparable
LiveCodeBench v6
GPT-5.6 Terra—
Kimi K2.585.0%
Source
Not directly comparable
SWE Multilingual
GPT-5.6 Terra—
Kimi K2.573%
Source
Not directly comparable
SWE-Rebench
GPT-5.6 Terra—
Kimi K2.558.5%
Source
Not directly comparable
React Native Evals
GPT-5.6 Terra—
Kimi K2.577.2%
Source
Not directly comparable
SciCode
GPT-5.6 Terra—
Kimi K2.548.7%
Source
Not directly comparable

Reasoning

ARC-AGI-2
GPT-5.6 Terra83.9%
Source
Kimi K2.5—
Not directly comparable
ARC-AGI-3
GPT-5.6 Terra0.8%
Source
Kimi K2.5—
Not directly comparable
LongBench v2
GPT-5.6 Terra—
Kimi K2.561%
Source
Not directly comparable

Multimodal

MMMU-Pro
GPT-5.6 Terra80.7%
Source
Kimi K2.578.5%
Source
GPT-5.6 Terra leads this result
MMMU-Pro w/ Python
GPT-5.6 Terra82%
Source
Kimi K2.5—
Not directly comparable
Video-MME
GPT-5.6 Terra—
Kimi K2.587.4%
Source
Not directly comparable
MMVU
GPT-5.6 Terra—
Kimi K2.580.4%
Source
Not directly comparable
VideoMMMU
GPT-5.6 Terra—
Kimi K2.586.6%
Source
Not directly comparable

Knowledge

GPQA
GPT-5.6 Terra92.9%
Source
Kimi K2.587.6%
Source
GPT-5.6 Terra leads this result
GPQA-D
GPT-5.6 Terra92.9%
Source
Kimi K2.587.6%
Source
GPT-5.6 Terra leads this result
HLE-Verified
GPT-5.6 Terra51.1%
Source
Kimi K2.5—
Not directly comparable
LABBench2
GPT-5.6 Terra81.2%
Source
Kimi K2.5—
Not directly comparable
HealthBench Professional
GPT-5.6 Terra57.7%
Source
Kimi K2.5—
Not directly comparable
HealthBench Hard
GPT-5.6 Terra32.7%
Source
Kimi K2.5—
Not directly comparable
GPQA Diamond (Vals)
GPT-5.6 Terra90.9%
Source
Kimi K2.5—
Not directly comparable
MMLU-Pro (Vals)
GPT-5.6 Terra86.7%
Source
Kimi K2.5—
Not directly comparable
SuperGPQA
GPT-5.6 Terra—
Kimi K2.569.2%
Source
Not directly comparable
MMLU-Pro
GPT-5.6 Terra—
Kimi K2.587.1%
Source
Not directly comparable
MMLU-Pro (Arcee)
GPT-5.6 Terra—
Kimi K2.587.1%
Source
Not directly comparable
HLE
GPT-5.6 Terra—
Kimi K2.530.1%
Source
Not directly comparable

Multilingual

MMLU-ProX
GPT-5.6 Terra—
Kimi K2.582.3%
Source
Not directly comparable
NOVA-63
GPT-5.6 Terra—
Kimi K2.556.0%
Source
Not directly comparable

Instruction following

IFEval
GPT-5.6 Terra—
Kimi K2.593.9%
Source
Not directly comparable

Math

FrontierMath (legacy)
GPT-5.6 Terra84.9%
Source
Kimi K2.5—
Not directly comparable
FrontierMath v2 (Tiers 1-3)
GPT-5.6 Terra84.900%
Source
Kimi K2.527.900%
Source
GPT-5.6 Terra leads this result
FrontierMath v2 (Tier 4)
GPT-5.6 Terra68.300%
Source
Kimi K2.54.200%
Source
GPT-5.6 Terra leads this result
AIME 2025
GPT-5.6 Terra—
Kimi K2.596.1%
Source
Not directly comparable
AIME26
GPT-5.6 Terra—
Kimi K2.595.8%
Source
Not directly comparable
AIME25 (Arcee)
GPT-5.6 Terra—
Kimi K2.596.3%
Source
Not directly comparable
HMMT Feb 2025
GPT-5.6 Terra—
Kimi K2.595.4%
Source
Not directly comparable
HMMT Nov 2025
GPT-5.6 Terra—
Kimi K2.591.1%
Source
Not directly comparable
HMMT Feb 2026
GPT-5.6 Terra—
Kimi K2.587.1%
Source
Not directly comparable
MMAnswerBench
GPT-5.6 Terra—
Kimi K2.581.8%
Source
Not directly comparable

69 public results · 8 shared

Related comparisons

Compare API pricing Read the methodology Open the model selector

Watch GPT-5.6 Terra vs Kimi K2.5

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

Read a sample issue

Join 2,000+ readers.

Last updated September 24, 2026

GPT-5.6 Terra vs Kimi K2.5

Decision reading

Which one for your work

Long documents

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

Coding work

Agentic work

Which one for a specific job

What is actually comparable

Shape of the matched evidence

Category results, on a stated basis

Multimodal

Agentic

Coding

Knowledge

Instruction following

Reasoning

Multilingual

Math

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

If you already use one of these models

Questions

Which is better, GPT-5.6 Terra or Kimi K2.5?

Which is better for coding, GPT-5.6 Terra or Kimi K2.5?

Which is better for agentic tasks, GPT-5.6 Terra or Kimi K2.5?

Which costs less, GPT-5.6 Terra or Kimi K2.5?

Which has the larger context window, GPT-5.6 Terra or Kimi K2.5?

Self-host vs API cost

Benchmark evidence

Agentic

Coding

Reasoning

Multimodal

Knowledge

Multilingual

Instruction following

Math

Related comparisons

Watch GPT-5.6 Terra vs Kimi K2.5