Data as of October 28, 2026 · How the score is built
Released Oct 28, 2026Google Gemini 4 Carbon evaluation methodology
Gemini 4 Carbon
Decision readingGemini 4 Carbon scores 88.4 out of 100 and ranks #1 of 215. This profile shows 10 source-displayable benchmark rows; its strongest eligible category is Coding at #3. API pricing is $2 input and $10 output per million tokens, with cached input at $0.1.
Released Oct 28, 2026 — see all recent releases
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
88.4/100
field median 50.1#1 of 215 ranked models
Public
#1of 215
Verified —
Price
$2input / $10 output
input median $0.75cached $0.10 · blended $6
Speed
100tok/s
field median 92 tok/sFirst token 97.48 s
Context
Not published
field median 256,000No exact-model context limit is sourced
Strongest published evidence
Coding ranks #3. Particularly well-suited for software development and code generation tasks.
Validate before choosing
10 published rows leave some tracked benchmark slots empty.
Source-linked · 10 displayable benchmark rows
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank Not rankedWeight 22%4 benchmarksVerified | 80.4 | 4 benchmarks | Verified | |||
| CodingRank #3 of 156Percentile 99thWeight 20%1 benchmarkVerified | 82.2 | 1 benchmark | Verified | |||
| ReasoningRank Not rankedWeight 17%1 benchmarkVerified | 79.5 | 1 benchmark | Verified | |||
| MultimodalWeight 12%2 benchmarksVerified | Score pending | 2 benchmarks | Verified | |||
| KnowledgeWeight 12%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingWeight 5%1 benchmarkVerified | Score pending | 1 benchmark | Verified | |||
| MathWeight 5%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured |
10 of 775 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic4/4 verified
- Coding1/1 verified
- Reasoning1/1 verified
- Multimodal2/2 verified
- KnowledgeNot measured
- MultilingualNot measured
- Inst. Following1/1 verified
- MathNot measured
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
Gemini 4 Carbon category percentile values
- AgenticNot eligible
- Coding99th percentile
- ReasoningNot eligible
- MultimodalNot eligible
- KnowledgeNot eligible
- MultilingualNot eligible
- Instruction followingNot eligible
- MathNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- AgenticNot ranked
- Coding#3/156
- ReasoningNot ranked
- MultimodalNot ranked
- KnowledgeNot ranked
- MultilingualNot ranked
- Inst. FollowingNot ranked
- MathNot ranked
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| FrontierSWE v2 | Score61.2% | Versus best verified row Best verified: GPT-6 Astra · 65.5% | Gap4.3 behind | Weight8% ref. weight | Benchmark exact |
Agentic4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| AutomationBench | Score53.9% | Versus best verified row Best verified: MiniMax M3.1 Flash · 62.2% | Gap8.3 behind | Weight6% ref. weight | Provider exact |
| OSWorld 2.1 OfflineOSWorld 2.1 offline subset (82 tasks) | Score81.5% | Versus best verified row Best verified: Claude Opus 5.5 · 87.2% | Gap5.7 behind | WeightDisplay only | Provider exact |
| CWE-bench v1 | Score71.0% | Versus best verified row Best verified: Gemini 4 Carbon · 71.0% | GapBest verified | WeightDisplay only | Benchmark exact |
| ExploitGym | Score53.5% | Versus best verified row Best verified: Gemini 4 Carbon · 53.5% | GapBest verified | WeightDisplay only | Provider exact |
Reasoning1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| GraphWalks BFS 256K–1M | Score96.2% | Versus best verified row Best verified: Gemini 4 Carbon · 96.2% | GapBest verified | WeightDisplay only | Provider exact |
Multimodal2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Chartography (no tools)Chartography without tools | Score72.5% | Versus best verified row Best verified: Gemini 4 Carbon · 72.5% | GapBest verified | WeightDisplay only | Provider exact |
| Blueprint-Bench 2 | Score58.6% | Versus best verified row Best verified: Gemini 4 Carbon · 58.6% | GapBest verified | WeightDisplay only | Provider exact |
Inst. Following1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Gray Swan IPI (15 attempts)Gray Swan Indirect Prompt Injection, attack success at 15 attempts | Score0.0% | Versus best verified row Best verified: Gemini 4 Carbon · 0.0% | GapBest verified | WeightDisplay only | Provider exact |
External signals1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SEC-Bench Pro | Score88.2% | Versus best verified row Best verified: Gemini 4 Carbon · 88.2% | GapBest verified | WeightDisplay only | Provider exact |
What it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
Gemini 4 Carbon · 88.4 score · $6 blended per million tokens
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Horizontal: blended price per million tokens, log scale · Vertical: public score
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Base entry
Gemini 4 Argon · 82.3Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not publishedGoogle Gemini 4 Carbon launch post
- Context window
- Not publishedGoogle Gemini 4 Carbon launch
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yetGoogle Gemini 4 Carbon launch
- Output modalities
- Not sourced yetGoogle Gemini 4 Carbon launch
- Parameters
- Not disclosed by the provider
- Availability
- Fairwind ProgramGoogle Gemini 4 Carbon launch
- Cloud regions
- Not tracked yet
- Lifecycle
- limited accessGoogle Gemini 4 Carbon launch
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.10 per million cached input tokensGoogle Gemini 4 Carbon launch post
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
Gemini 4 Carbon ranks #1 of 215 on the public leaderboard with a score of 88.42/100. It does not yet have enough sourced coverage for a verified position.
Gemini 4 Carbon is a proprietary model. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Google is giving Gemini 4 Carbon first to a subset of cyber defenders, then to more trusted partners through its Fairwind Program, before a wider release to developers, businesses, and consumers. Google has not yet published a Gemini API model ID, context window, or model documentation for Carbon.
Gemini 4 Carbon sits in the Gemini 4 family with Gemini 4 Argon. Its explicit predecessor is Gemini 4 Argon. 10 of 775 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Coding at #3. particularly well-suited for software development and code generation tasks.
Last updated October 28, 2026. Runtime fields remain blank until a sourced snapshot exists.
Questions
How does Gemini 4 Carbon perform overall in AI benchmarks?
Gemini 4 Carbon ranks #1 out of 215 models on the public BenchAlign leaderboard, with a score of 88.42/100. Its evidence status is Estimated, and this profile shows 10 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is Gemini 4 Carbon good for coding and programming?
Gemini 4 Carbon ranks #3 out of 156 eligible models for coding and programming, with a public category score of 82.2/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Gemini 4 Carbon good for reasoning and logic?
Gemini 4 Carbon has source-displayable benchmark coverage for reasoning and logic, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Gemini 4 Carbon good for agentic tool use and computer tasks?
Gemini 4 Carbon has source-displayable benchmark coverage for agentic tool use and computer tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Gemini 4 Carbon good for multimodal and grounded tasks?
Gemini 4 Carbon has source-displayable benchmark coverage for multimodal and grounded tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Gemini 4 Carbon good for instruction following?
Gemini 4 Carbon has source-displayable benchmark coverage for instruction following, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Which sibling models are related to Gemini 4 Carbon?
Gemini 4 Carbon belongs to the Gemini 4 family. Related tracked variants include Gemini 4 Argon. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.
Does Gemini 4 Carbon have full benchmark coverage on BenchLM?
No. Gemini 4 Carbon currently has 23 source-displayable rows. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Gemini 4 Carbon?
Gemini 4 Carbon's provider has not published a context-window limit for this exact model. The profile leaves the value unavailable instead of borrowing a number from an earlier family member or a third-party route. Maximum output length is tracked separately when the provider documents it.