Gemini API Family
Use one EvoLink API to access all Gemini models. Compare Gemini 3.1 Pro, 3 Flash, 3.1 Flash Lite, 2.5 Pro, 2.5 Flash, and 2.5 Flash Lite on pricing, context window, modality, and reasoning fit — then pick the right route for your workload.
6 routes
Pro, Flash, and Lite tiers for every budget
Unified API access
OpenAI compatible, one key for all Gemini
Choose by workflow
Match Pro vs Flash vs Lite to your task
Compare Gemini API routes
Start from the workload: flagship reasoning, production Flash traffic, low-cost extraction, or long-context multimodal analysis.
| Route | Best for | Pricing | Context window | Modality | Status |
|---|---|---|---|---|---|
Gemini 3.1 Pro Preview Flagship reasoning | Highest-quality Gemini reasoning, coding, agents, and long-context analysis. | $2/$12 <=200K; $4/$18 >200K | 1M input / 64K output | Text, code, image, video, audio, PDF inputs | Preview flagship |
Gemini 3 Flash Preview Fast Gemini 3 | Low-latency multimodal apps that need stronger Gemini 3 behavior than older Flash routes. | $0.50/$3.00 per MTok (audio in: $1.00) | 1M input / 64K output | Text, image, video, audio, PDF inputs | Preview route |
Gemini 3.1 Flash Lite Preview Cheapest Gemini 3 | High-volume translation, classification, extraction, and batch text workloads at the lowest Gemini 3.x cost. | $0.25/$1.50 per MTok (audio in: $0.50) | 1M input / 64K output | Text, image, video, audio, PDF inputs | Preview route |
Gemini 2.5 Pro Stable Pro | Production reasoning, coding help, analysis, and complex multimodal tasks. | $1.25/$10 <=200K; $2.50/$15 >200K | 1M input / 64K output | Text, image, video, audio, PDF inputs | Stable deep reasoning |
Gemini 2.5 Flash Production Flash | Fast chat, extraction, summaries, and multimodal production traffic. | $0.30/$2.50 per MTok (audio in: $1.00) | 1M input / 64K output | Text, image, video, audio, PDF inputs | Production workhorse |
Gemini 2.5 Flash Lite Lowest cost | High-volume classification, extraction, routing, and lightweight chat flows. | $0.10/$0.40 per MTok (audio in: $0.30) | 1M input / 64K output | Text, audio inputs | Lowest-cost text route |
How to decide which Gemini model to use
Follow these 4 rules to narrow down your choice across Pro, Flash, and Lite tiers.
Start with reasoning depth
Complex coding agents, multi-step tool use, deep document analysis, and high-accuracy output — start with Gemini 3.1 Pro or Gemini 2.5 Pro.
Then check latency and throughput needs
Production chat, support bots, real-time extraction, and high-frequency multimodal apps — compare Gemini 3 Flash or Gemini 2.5 Flash.
Then check cost sensitivity
High-volume classification, batch text processing, routing, and lightweight extraction — compare Gemini 3.1 Flash Lite or Gemini 2.5 Flash Lite.
Finally, consider mixed-complexity workflows
If the same pipeline mixes simple classification with deep reasoning steps, consider EvoLink Smart Router instead of hardcoding one Gemini model.
Smart Router →If you already know your task type, find the recommended starting point in the table below.
Choose a Gemini model by workflow: reasoning, speed, cost, and multimodal tasks
Match your primary task to the right Gemini route.
| Your task | Recommended start | Good fit if... | Watch out for |
|---|---|---|---|
| Complex reasoning and coding agents | Gemini 3.1 Pro | You need highest-quality Gemini reasoning, multi-step tool use, or deep code analysis | Higher cost — use Flash for simpler tasks |
| Stable deep reasoning with multimodal | Gemini 2.5 Pro | You need production-grade reasoning with broad multimodal support and proven stability | Slightly lower capability ceiling than 3.1 Pro |
| Low-latency multimodal apps | Gemini 3 Flash | You need fast responses with Gemini 3 generation capabilities across text, image, audio, and video | Preview route — check stability requirements |
| Production chat and extraction | Gemini 2.5 Flash | You need a proven production workhorse for chat, summaries, extraction at scale | Good default for most production workloads |
| High-volume batch text at lowest cost | Gemini 2.5 Flash Lite | Tasks are classification, routing, or short responses where cost matters most | Limited to text and audio input only |
| Mixed-complexity text workflows | EvoLink Smart Router | Same pipeline has both simple and complex tasks across Gemini and other providers | Best when you don't want manual model routing logic |
Gemini API workflows: agents, chat, documents, and multimodal processing
See how Gemini models fit into real products, agents, and content processing pipelines.
Reasoning and coding agents
For code generation, bug fixing, multi-step tool use, and complex analysis agents. If output quality directly affects product behavior, start with Gemini 3.1 Pro. For proven stability, compare Gemini 2.5 Pro.
Production chat and support
For support bots, in-app assistants, knowledge base Q&A, and high-frequency multi-turn conversations. Test with Gemini 2.5 Flash first for proven throughput, then compare Flash Lite for lower cost.
Long document and multimodal analysis
For PDF analysis, video understanding, audio transcription, and multi-file research workflows. Gemini's 1M context window and native multimodal support make Pro and Flash routes strong choices.
Agent routing and mixed tasks
For workflows where classification, extraction, reasoning, and generation coexist in the same pipeline. Use EvoLink Smart Router to automatically route between Gemini and other providers via evolink/auto.
View Gemini model details
Use this page to compare, then visit individual model pages for pricing details, playground access, and integration guides.
Gemini 3.1 Pro Preview
Flagship reasoning
- Context
- 1M input / 64K output
- Pricing
- $2/$12 <=200K; $4/$18 >200K
Gemini 3 Flash Preview
Fast Gemini 3
- Context
- 1M input / 64K output
- Pricing
- $0.50/$3.00 per MTok (audio in: $1.00)
Gemini 3.1 Flash Lite Preview
Cheapest Gemini 3
- Context
- 1M input / 64K output
- Pricing
- $0.25/$1.50 per MTok (audio in: $0.50)
Gemini 2.5 Pro
Stable Pro
- Context
- 1M input / 64K output
- Pricing
- $1.25/$10 <=200K; $2.50/$15 >200K
Gemini 2.5 Flash
Production Flash
- Context
- 1M input / 64K output
- Pricing
- $0.30/$2.50 per MTok (audio in: $1.00)
Gemini 2.5 Flash Lite
Lowest cost
- Context
- 1M input / 64K output
- Pricing
- $0.10/$0.40 per MTok (audio in: $0.30)
Access all Gemini models through one EvoLink API
All 6 Gemini routes are available through a single EvoLink API key and OpenAI-compatible endpoint. Switch between Pro, Flash, and Lite by changing the model parameter — no separate accounts or keys needed.
Switch model="gemini-3.1-pro" to model="gemini-2.5-flash" without rebuilding your integration.How to think about Gemini API cost: Pro vs Flash vs Lite
Pro routes: reasoning justifies the premium
Gemini 3.1 Pro and 2.5 Pro cost more per token, but complex coding agents, deep document analysis, and multi-step tool use produce higher-value outputs. Don't default to Pro for simple extraction or classification.
Flash routes: best balance for production volume
Gemini 3 Flash and 2.5 Flash deliver strong multimodal capabilities at a fraction of Pro pricing. Start here for chat, summaries, and production-scale extraction before considering Pro.
Lite routes: minimize cost for simple high-volume tasks
Gemini 3.1 Flash Lite and 2.5 Flash Lite offer the lowest per-token cost. Use them for classification, routing, batch text, and short responses where reasoning depth is not critical.
Pricing summary
Gemini routes range from $0.10/MTok input (Flash Lite) to $4.00/MTok input (Pro >200K). All use per-token pricing via EvoLink.
Gemini 3.1 Pro
$2/$12 — $4/$18 /MTok
Context: 1M
Flagship reasoning with 1M context. Tiered pricing: $2/$12 under 200K, $4/$18 over 200K input tokens.
Gemini 3 Flash
$0.50/$3.00 /MTok
Context: 1M
Gemini 3 generation Flash route at $0.50/$3.00 per MTok with 1M context.
Gemini 3.1 Flash Lite
$0.25/$1.50 /MTok
Context: 1M
Cheapest Gemini 3 route at $0.25/$1.50 per MTok for batch text workloads.
Gemini 2.5 Pro
$1.25/$10 — $2.50/$15 /MTok
Context: 1M
Stable deep reasoning at $1.25/$10 under 200K, $2.50/$15 over 200K.
Gemini 2.5 Flash
$0.30/$2.50 /MTok
Context: 1M
Production workhorse at $0.30/$2.50 per MTok with full multimodal support.
Gemini 2.5 Flash Lite
$0.10/$0.40 /MTok
Context: 1M
Lowest-cost Gemini route at $0.10/$0.40 per MTok for text and audio.
Gemini guides and comparisons
Use these guides when you need more context before choosing a route.
Gemini 3.1 Pro vs GPT-5.2 vs Claude Opus
Compare flagship models for reasoning, coding, and production agent workloads.
GPT-5.4 vs Claude Opus 4.6 vs Gemini 3.1 Pro
See how the newest premium models compare on practical API selection.
Gemini 3 Pro deprecation migration guide
Move old Gemini 3 Pro Preview traffic to current Gemini routes without breaking production behavior.
OpenCode integration with Gemini routes
See how to access Gemini alongside Claude and GPT models through EvoLink's unified API layer.
Gemini API FAQ
Everything you need to know about the product and billing.