Google: Models Intelligence, Performance & Price
This analysis is intended to support you in choosing the best model provided by Google for your use-case.
Most Intelligent
Intelligence Index
Total 57 models
Lowest Time per Task
Time per Intelligence Index task (minutes)
Total 57 models
Lowest Price
Blended price (per 1M tokens)
Total 57 models
Google offers 57 models, each with different intelligence, performance, and pricing characteristics. Below is a comparison of the key metrics across models.
- For intelligence, the top models on Google are Gemini 4 Carbon (high) (60), Claude Opus 5.5 (max) (59), and Gemini 4 Carbon (medium) (58).
- For time per task, Gemini 3.1 Flash-Lite (AI Studio) (0.8 min), Gemini 4 Hydrogen (low) (1.5 min), and Gemini 3.5 Flash-Lite AI Studio (1.6 min) finish an Intelligence Index task fastest on Google. Time per task varies significantly across models, with the slowest taking 3.8x longer than the quickest.
- For latency, Gemini 2.5 Flash-Lite (non-reasoning) (AI Studio) (0.35s), Claude 4.5 Haiku (non-reasoning) Vertex (0.60s), and Gemini 2.5 Flash (non-reasoning) (AI Studio) (0.60s) offer the lowest time to first answer token.
- For pricing, Gemma 4 31B (AI Studio) ($0.00), Gemma 4 26B A4B AI Studio ($0.00), and Gemini 2.5 Flash-Lite (non-reasoning) (AI Studio) ($0.07) offer the lowest blended prices per 1M tokens.
- For context window size, Llama 4 Scout Vertex (1M), Gemini 4 Carbon (high) (1M), and Claude Opus 5.5 (max) (1M) support the largest context windows on Google.
Intelligence
Artificial Analysis Intelligence Index
Evaluations
Agentic knowledge work, Elo
Knowledge
1 - hallucination rate
Agentic SaaS workflows
Agentic scientific research workflows in a terminal
Agentic software engineering in open-source repositories
Reasoning & knowledge
Professional document reasoning, All-pass
Agentic coding & terminal use
Physics reasoning
Quantitative analysis on spreadsheets & documents
Legal agentic work, Hallucination-Gated All-Pass Rate
Agentic business operations
Agentic real-world work tasks, Elo
Agentic interactive 3D experiences, Elo
Product and brand videos designed in code, Elo
Kubernetes incident root-cause analysis
Medical long context reasoning
Visual reasoning
Rich, varied, well-crafted language
Well-structured, scannable, answer-first
Truthful, acknowledges uncertainty, non-sycophantic
Use of filler and generic language
Intelligence Index vs. Price
Pricing
Cost per Intelligence Index Task
Intelligence Index vs. Price
Speed
Average of the single-request and multi-turn agentic workloads measured
Time per Intelligence Index Task
Quality
Endpoint Accuracy Index
Latency
Average of the single-request and multi-turn agentic workloads measured
Time to First Answer Token
Reliability
Share of requests the endpoint failed, across all the single-request and multi-turn agentic workloads measured
Endpoint Error Rate
Caching
Cache Hit Rate
Cost per Task vs. Cache Hit Rate
Further Analysis | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Gemini 4 Carbon (high) | 60 | $5.48 | 32.8 | 72.65 | 11.9% | 1M | Proprietary | |||
Claude Opus 5.5 (max) | 59 | $21.07 | 24.6 | 342.69 | 4.7% | 1M | Proprietary | |||
Gemini 4 Carbon (medium) | 58 | $4.68 | 28.9 | 21.52 | 7.0% | 1M | Proprietary | |||
Claude Sonnet 5.5 (max) | 55 | $14.47 | 30.3 | 259.97 | 2.4% | 1M | Proprietary | |||
Gemini 4 Carbon (low) | 54 | $5.22 | 27.3 | 10.41 | 3.0% | 1M | Proprietary | |||
Gemini 4 Hydrogen (xhigh) | 54 | $3.16 | 14.9 | 42.99 | 5.2% | 1M | Proprietary | |||
Gemini 4 Hydrogen (high) | 51 | $1.87 | 8.2 | 15.99 | 3.0% | 1M | Proprietary | |||
Gemini 4 Argon (high) AI Studio | 50 | $4.97 | 29.6 | 45.59 | 4.3% | 1M | Proprietary | |||
Gemini 4 Argon (high) (Vertex) | 50 | $3.84 | 36.8 | 47.98 | 9.4% | 1M | Proprietary | |||
Gemini 4 Argon (medium) AI Studio | 48 | $4.31 | 28.1 | 12.24 | 2.4% | 1M | Proprietary | |||
Gemini 4 Argon (medium) (Vertex) | 48 | $4.31 | 34.4 | 9.07 | 4.8% | 1M | Proprietary | |||
Claude Opus 5 (max) | 44 | $15.29 | 23.2 | 30.31 | 0.0% | 1M | Proprietary | |||
Gemini 4 Hydrogen (medium) | 41 | $0.80 | 3.9 | 12.07 | 3.0% | 1M | Proprietary | |||
Claude Opus 5.5 (low) | 35 | $2.64 | 3.0 | 4.55 | 0.0% | 1M | Proprietary | |||
Gemini 3.8 Flash (high) AI Studio | 31 | $7.77 | 14.2 | 19.79 | 0.7% | 1M | Proprietary | |||
Gemini 3.8 Flash (medium) AI Studio | 30 | $6.26 | 11.6 | 13.95 | 0.0% | 1M | Proprietary | |||
Gemini 4 Hydrogen (low) | 27 | $0.29 | 1.4 | 5.95 | 4.2% | 1M | Proprietary | |||
Claude Sonnet 5.5 (low) | 26 | $1.03 | 2.4 | 1.99 | 0.0% | 1M | Proprietary | |||
Gemini 3.8 Flash (low) AI Studio | 23 | -- | -- | 4.26 | 0.0% | 1M | Proprietary | |||
Gemini 3.5 Flash-Lite AI Studio | 12 | $0.69 | 1.6 | 4.82 | 2.3% | 1M | Proprietary | |||
Claude 4.5 Haiku Vertex | 7 | $0.57 | 3.5 | 16.85 | 0.0% | 200k | Proprietary | |||
Gemini 3.1 Flash-Lite (AI Studio) | 6 | $0.19 | 0.8 | 8.26 | 0.0% | 1M | Proprietary | |||
Gemma 4 31B (AI Studio) | 4 | -- | -- | 52.52 | 0.0% | 262k | Open | |||
gpt-oss-120b (high) Vertex | 4 | $0.21 | 9.3 | 26.84 | 8.2% | 131k | Open | |||
Gemini 3.5 Flash (high) AI Studio | -- | -- | -- | 9.99 | 0.0% | 1M | Proprietary | |||
Gemini 3.5 Flash (minimal) AI Studio | -- | -- | -- | 2.62 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Pro (May) (AI Studio) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Claude Opus 5 (xhigh) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Gemini 2.5 Flash-Lite (non-reasoning) (AI Studio) | -- | -- | -- | 0.36 | 2.0% | 1M | Proprietary | |||
Claude Opus 5 (low) | -- | -- | -- | 2.07 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Flash (AI Studio) | -- | -- | -- | 12.98 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Flash (Vertex) | -- | -- | -- | 15.01 | 0.0% | 1M | Proprietary | |||
Gemma 3 4B (AI Studio) | -- | -- | -- | -- | -- | 128k | Open | |||
Claude Sonnet 5 (max) | -- | -- | -- | 135.62 | 0.0% | 1M | Proprietary | |||
Claude Opus 4.5 Vertex | -- | -- | -- | 9.78 | 0.0% | 200k | Proprietary | |||
Gemini 2.5 Pro (AI Studio) | -- | -- | -- | 21.94 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Pro Vertex | -- | -- | -- | 21.21 | 0.0% | 1M | Proprietary | |||
Gemma 3 27B (AI Studio) | -- | -- | -- | -- | -- | 128k | Open | |||
Claude Sonnet 4.6 (non-reasoning, high) | -- | -- | -- | 1.23 | 0.0% | 200k | Proprietary | |||
Gemma 3 12B (AI Studio) | -- | -- | -- | -- | -- | 128k | Open | |||
Claude Opus 4.5 (non-reasoning) Vertex | -- | -- | -- | 0.94 | 0.0% | 200k | Proprietary | |||
Claude Opus 4.6 (non-reasoning, high) | -- | -- | -- | 1.71 | 0.0% | 1M | Proprietary | |||
Gemini 3.5 Flash (medium) | -- | -- | -- | 8.50 | 0.0% | 1M | Proprietary | |||
Claude 4 Opus (non-reasoning) Vertex | -- | -- | -- | -- | -- | 200k | Proprietary | |||
Claude Opus 4.6 (max) | -- | -- | -- | 24.54 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Flash (non-reasoning) (Vertex) | -- | -- | -- | 0.90 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Flash (non-reasoning) (AI Studio) | -- | -- | -- | 0.67 | 4.4% | 1M | Proprietary | |||
Gemma 4 26B A4B AI Studio | -- | -- | -- | 27.21 | 1.7% | 262k | Open | |||
Claude Opus 5 (high) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Gemini 2.5 Flash-Lite (Sep) (non-reasoning) (AI Studio) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Gemini 3.6 Flash (high) AI Studio | -- | -- | -- | 10.40 | 2.8% | 1M | Proprietary | |||
Gemini 3 Flash (non-reasoning) (AI Studio) | -- | -- | -- | 1.22 | 0.0% | 1M | Proprietary | |||
Gemini 3 Flash (AI Studio) | -- | -- | -- | 4.70 | 0.0% | 1M | Proprietary | |||
Gemini 2.5 Flash-Lite (Sep) (AI Studio) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Claude 4 Opus Vertex | -- | -- | -- | -- | -- | 200k | Proprietary | |||
Claude 4.5 Sonnet Vertex | -- | -- | -- | 10.57 | 0.0% | 1M | Proprietary | |||
Claude 4.5 Sonnet (non-reasoning) Vertex | -- | -- | -- | 1.11 | 0.0% | 1M | Proprietary | |||
Claude Opus 4.8 (max) | -- | -- | -- | 35.97 | 0.0% | 1M | Proprietary | |||
Gemma 3n E2B (AI Studio) | -- | -- | -- | -- | -- | 32k | Open | |||
Llama 4 Scout Vertex | -- | -- | -- | 0.92 | 0.7% | 1M | Open | |||
Gemini 3 Pro Preview (low) (Vertex) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Gemini 3.1 Pro Preview (Vertex) | -- | -- | -- | 19.48 | 0.9% | 1M | Proprietary | |||
Gemini 3.1 Pro Preview (AI Studio) | -- | -- | -- | 14.94 | 0.8% | 1M | Proprietary | |||
Claude 4.5 Haiku (non-reasoning) Vertex | -- | -- | -- | 0.60 | 0.0% | 200k | Proprietary | |||
gpt-oss-120b (low) Vertex | -- | -- | -- | 3.22 | 3.4% | 131k | Open | |||
Claude 4.1 Opus Vertex | -- | -- | -- | -- | -- | 200k | Proprietary | |||
Claude Fable 5 | -- | -- | -- | 65.02 | 0.0% | 1M | Proprietary | |||
Gemini 3 Pro Preview (high) (Vertex) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Gemma 3 1B (AI Studio) | -- | -- | -- | -- | -- | 32k | Open | |||
Claude 3.7 Sonnet (non-reasoning) Vertex | -- | -- | -- | -- | -- | 200k | Proprietary | |||
Claude Sonnet 4.6 (max) | -- | -- | -- | 106.15 | 5.4% | 200k | Proprietary | |||
Claude Opus 4.7 (max) | -- | -- | -- | 12.70 | 0.0% | 1M | Proprietary | |||
Claude 4.1 Opus (non-reasoning) Vertex | -- | -- | -- | -- | -- | 200k | Proprietary | |||
Gemini 2.5 Flash-Lite (AI Studio) | -- | -- | -- | 18.76 | 4.0% | 1M | Proprietary | |||
Gemini 2.0 Flash (exp) (AI Studio) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Claude Opus 5 (medium) | -- | -- | -- | -- | -- | 1M | Proprietary | |||
Key definitions
Frequently Asked Questions
Common questions about Google
Google offers 57 models that we track: Gemini 4 Carbon (high), Claude Opus 5.5 (max), Gemini 4 Carbon (medium), Claude Sonnet 5.5 (max), Gemini 4 Carbon (low), Gemini 4 Hydrogen (xhigh), Gemini 4 Hydrogen (high), Gemini 4 Argon (high) AI Studio, Gemini 4 Argon (high) (Vertex), Gemini 4 Argon (medium) AI Studio, Gemini 4 Argon (medium) (Vertex), Claude Opus 5 (max), Gemini 4 Hydrogen (medium), Claude Opus 5.5 (low), Gemini 3.8 Flash (high) AI Studio, Gemini 3.8 Flash (medium) AI Studio, Gemini 4 Hydrogen (low), Claude Sonnet 5.5 (low), Gemini 3.8 Flash (low) AI Studio, Gemini 3.5 Flash-Lite AI Studio, Claude 4.5 Haiku Vertex, Gemini 3.1 Flash-Lite (AI Studio), Gemma 4 31B (AI Studio), gpt-oss-120b (high) Vertex, Gemini 3.5 Flash (high) AI Studio, Gemini 3.5 Flash (minimal) AI Studio, Gemini 2.5 Flash-Lite (non-reasoning) (AI Studio), Claude Opus 5 (low), Gemini 2.5 Flash (AI Studio), Gemini 2.5 Flash (Vertex), Claude Sonnet 5 (max), Claude Opus 4.5 Vertex, Gemini 2.5 Pro (AI Studio), Gemini 2.5 Pro Vertex, Claude Sonnet 4.6 (non-reasoning, high), Claude Opus 4.5 (non-reasoning) Vertex, Claude Opus 4.6 (non-reasoning, high), Gemini 3.5 Flash (medium), Claude Opus 4.6 (max), Gemini 2.5 Flash (non-reasoning) (Vertex), Gemini 2.5 Flash (non-reasoning) (AI Studio), Gemma 4 26B A4B AI Studio, Gemini 3.6 Flash (high) AI Studio, Gemini 3 Flash (non-reasoning) (AI Studio), Gemini 3 Flash (AI Studio), Claude 4.5 Sonnet Vertex, Claude 4.5 Sonnet (non-reasoning) Vertex, Claude Opus 4.8 (max), Llama 4 Scout Vertex, Gemini 3.1 Pro Preview (Vertex), Gemini 3.1 Pro Preview (AI Studio), Claude 4.5 Haiku (non-reasoning) Vertex, gpt-oss-120b (low) Vertex, Claude Fable 5, Claude Sonnet 4.6 (max), Claude Opus 4.7 (max), and Gemini 2.5 Flash-Lite (AI Studio).
The most intelligent model available on Google is Gemini 4 Carbon (high) with an Intelligence Index score of 60.
The model with the lowest time per task on Google is Gemini 3.1 Flash-Lite (AI Studio) at 0.8 minutes per Intelligence Index task. Time per task is the output tokens per task divided by end-to-end output speed, so lower means a task finishes sooner.
The model with the lowest time to first answer token on Google is Gemini 2.5 Flash-Lite (non-reasoning) (AI Studio) at 0.35s. Lower latency means faster initial response time.
The most affordable model on Google by blended price is Gemma 4 31B (AI Studio) at $0.00 per 1M tokens (7:2:1 cache hit/input/output ratio).
Prices on Google range from $0.00 per 1M tokens for Gemma 4 31B (AI Studio) to $7.70 per 1M tokens for Claude Fable 5.
Yes, Google offers an OpenAI-compatible API, making it easy to switch from OpenAI or use existing OpenAI SDK integrations.
43 of 57 models on Google support JSON mode for structured output.
Yes, all 57 models on Google support function calling (tool use).
Yes, Google offers 46 reasoning models: Gemini 4 Carbon (high), Claude Opus 5.5 (max), Gemini 4 Carbon (medium), Claude Sonnet 5.5 (max), Gemini 4 Carbon (low), Gemini 4 Hydrogen (xhigh), Gemini 4 Hydrogen (high), Gemini 4 Argon (high) AI Studio, Gemini 4 Argon (high) (Vertex), Gemini 4 Argon (medium) AI Studio, Gemini 4 Argon (medium) (Vertex), Claude Opus 5 (max), Gemini 4 Hydrogen (medium), Claude Opus 5.5 (low), Gemini 3.8 Flash (high) AI Studio, Gemini 3.8 Flash (medium) AI Studio, Gemini 4 Hydrogen (low), Claude Sonnet 5.5 (low), Gemini 3.8 Flash (low) AI Studio, Gemini 3.5 Flash-Lite AI Studio, Claude 4.5 Haiku Vertex, Gemini 3.1 Flash-Lite (AI Studio), Gemma 4 31B (AI Studio), gpt-oss-120b (high) Vertex, Gemini 3.5 Flash (high) AI Studio, Claude Opus 5 (low), Gemini 2.5 Flash (AI Studio), Gemini 2.5 Flash (Vertex), Claude Sonnet 5 (max), Claude Opus 4.5 Vertex, Gemini 2.5 Pro (AI Studio), Gemini 2.5 Pro Vertex, Gemini 3.5 Flash (medium), Claude Opus 4.6 (max), Gemma 4 26B A4B AI Studio, Gemini 3.6 Flash (high) AI Studio, Gemini 3 Flash (AI Studio), Claude 4.5 Sonnet Vertex, Claude Opus 4.8 (max), Gemini 3.1 Pro Preview (Vertex), Gemini 3.1 Pro Preview (AI Studio), gpt-oss-120b (low) Vertex, Claude Fable 5, Claude Sonnet 4.6 (max), Claude Opus 4.7 (max), and Gemini 2.5 Flash-Lite (AI Studio). Reasoning models use extended thinking to work through complex problems before providing an answer.
Yes, 5 of 57 models on Google are open weight models: Gemma 4 31B (AI Studio), gpt-oss-120b (high) Vertex, Gemma 4 26B A4B AI Studio, Llama 4 Scout Vertex, and gpt-oss-120b (low) Vertex.
Yes, provider performance can vary over time due to infrastructure changes, load balancing, and updates. We continuously benchmark all providers and display historical performance trends in the "Over Time" charts.
When choosing a model on Google, consider: intelligence (for quality-sensitive tasks), time per task (how quickly the model completes an Intelligence Index task), latency (for interactive applications requiring quick first responses), pricing (for cost-sensitive workloads), and features like context window size, JSON mode, or function calling support.