Global AI Token Comparison
Global AI Token Plan comparison home

Openrouter Token Plan Pricing, Top-up, Deals & Official Entry

OpenRouter Token Plan pricing comparison is about prepaid Credits quota and billing rules, not a traditional monthly package. API and OpenRouter Chat share USD Credits. This page highlights the latest multi-type models from a global TOP15 set of providers (OpenAI, Anthropic, Google, DeepSeek, xAI, Qwen, Z.ai, MiniMax, Kimi, Xiaomi, Meta, Mistral, Perplexity, NVIDIA, Tencent) while the full catalog still covers 400+ models at openrouter.ai/models pass-through prices with no inference-token markup and a 5.5% top-up fee ($0.80 minimum). Vendor tables, free-model 50→1000/day after $10 credits, AliPay / USDC, BYOK, and Claude Code / Cursor / OpenAI SDK setup help you decide whether Credits fit as a unified multi-model balance.

Token PlanTOP15 vendors · Credits no markupCredits Prepaid
SubscriptionToken API

Openrouter Token Plan Latest Updates

Analysis

OpenRouter Pricing Comparison: Which of 400+ Models Is Cheapest? Per-Provider Rates, 5.5% Top-Up Fee, and Is It Worth It

OpenRouter has no monthly plan—only prepaid Credits with pay-as-you-go billing. Its 400+ models bill at provider pass-through rates with no markup, but top-ups carry a 5.5% fee ($0.80 minimum). The cheapest paid model, Qwen Flash, starts at $0.03/1M input tokens, while GPT-5.6 Sol tops out at $30/1M output—nearly a 50x gap between flagships. This article tallies entry and flagship prices across 15 providers, free-model limits (50→1000/day), and the direct-API vs OpenRouter math so you can decide whether it's worth it.

View all updates

Below is the complete pricing comparison for Openrouter Token Plan, covering 15 plan tiers, core model capabilities, quotas, and official entry. All data is sourced from the official website to help you decide whether it’s worth it.

Last updated：

Openrouter Token Plan Price Comparison: Core Models

GPT-5.6 SolGPT-5.6 TerraGPT-5.6 LunaGPT-5.3-CodexGPT-5.4 Image 2Claude Opus 5Claude Sonnet 5Claude Fable 5Claude Haiku 4.5Claude Opus 5 FastGemini 3.6 FlashGemini 3.5 FlashGemini 3.5 Flash LiteGemini 3.1 Pro PreviewGemini 3 Pro ImageDeepSeek V4 Pro 0813DeepSeek V4 Flash 0731DeepSeek V4 ProDeepSeek V4 Flash 0423Grok 4.6Grok 4.5Grok Build 0.1Grok 4.3Grok 4.20 Multi-AgentQwen3.8 MaxQwen3.8 2.4T A95BQwen3.7 MaxQwen3.7 PlusQwen3.7 FlashGLM 5.2GLM 5.1GLM 5V TurboGLM 5 TurboGLM 4.7 FlashMiniMax M3MiniMax M2.7MiniMax M2.5MiniMax M2.1Kimi K3Kimi K2.7 CodeKimi K2.6Kimi K2.5MiMo-V2.5-ProMiMo-V2.5Llama 4 MaverickLlama 4 ScoutLlama 3.3 70BLlama Guard 4 12BMistral Medium 3.5Mistral Large 3Mistral Small 4Ministral 14BMinistral 8BSonar Pro SearchSonar Reasoning ProSonar ProSonar Deep ResearchSonarNemotron 3.5 LightningNemotron 3 UltraNemotron 3 SuperNemotron 3 NanoHy3Hy3 previewHunyuan A13B
GPT-5.6 Sol

OpenAI GPT-5.6 flagship, ~1.05M context at $5/M input and $30/M output. Suited for complex reasoning, multi-step coding, and agent workflows; model id: `openai/gpt-5.6-sol`.

GPT-5.6 Terra

Balanced GPT-5.6 tier, ~1.05M context at $1/M input and $6/M output. Capability and cost between Sol and Luna for everyday professional work; model id: `openai/gpt-5.6-terra`.

GPT-5.6 Luna

Efficient GPT-5.6 tier, ~1.05M context at $0.10/M input and $0.60/M output. Built for high-volume, low-latency chat, classification, and light agents; model id: `openai/gpt-5.6-luna`.

GPT-5.3-Codex

OpenAI coding-focused model, 400K context at $1.75/M input and $14/M output. Suited to CLI, repo-scale edits, and agent coding; model id: `openai/gpt-5.3-codex`.

GPT-5.4 Image 2

OpenAI image model, 272K context at $8/M input and $15/M output. For image generation and understanding workloads; model id: `openai/gpt-5.4-image-2`.

Openrouter Token Plan Price Comparison: Plans

OpenAI
GPT-5.6 lineup · Codex · imageRecommended
Input
$0.10
Output
$0.60
Luna entry · flagship Sol $5/$30 · 5.5% top-up fee
Usage
Call the full OpenAI lineup via prepaid OpenRouter Credits with openai/* model ids at openrouter.ai/models pass-through rates and no inference-token markup; top-ups carry a 5.5% fee ($0.80 minimum).
Models
Core models: GPT-5.6 Sol flagship reasoning/agent (~1.05M context, $5/$30), Terra balanced daily ($1/$6), Luna high-throughput ($0.10/$0.60), GPT-5.3-Codex coding ($1.75/$14), GPT-5.4 Image 2 ($8/$15).
Highlights
Suited for complex reasoning, multi-step agents, repo-scale coding, and multimodal image work; one OpenRouter API key switches Sol/Terra/Luna/Codex/Image with provider routing and fallback.
Anthropic
Claude 5 series
Input
$1
Output
$5
Haiku entry · Opus $5/$25
Usage
Call the full Anthropic Claude lineup via OpenRouter Credits with anthropic/* model ids at models-page pass-through rates—ideal for long-horizon agents, code review, and professional writing without a separate Claude bill.
Models
Core models: Claude Opus 5 flagship (~1M context, $5/$25), Sonnet 5 daily driver ($2/$10), Fable 5 autonomous knowledge work ($10/$50), Haiku 4.5 light/fast ($1/$5), Opus 5 Fast ($10/$50).
Highlights
Strong coding, vision, and multi-step tool use; Fast variants trade ~2× price for higher output speed on latency-sensitive production paths.
Google
Gemini 3.x
Input
$0.30
Output
$2.50
Flash Lite entry · 3.6 Flash $1.5/$7.5
Usage
Call the full Google Gemini lineup via OpenRouter Credits with google/* model ids at models-page pass-through rates, supporting text/image/audio/video multimodal input and ~1M long context.
Models
Core models: Gemini 3.6 Flash efficient coding ($1.5/$7.5), 3.5 Flash ($1.5/$9), 3.5 Flash Lite ($0.3/$2.5), 3.1 Pro Preview flagship ($2/$12), 3 Pro Image ($2/$12).
Highlights
Suited for web/app development, agent subtasks, high-throughput batching, and image work; one key switches among Flash/Pro/Image tiers.
DeepSeek
V4 Pro / Flash
Input
$0.08
Output
$0.18
Flash 0731 entry · Pro 0813 $0.435/$0.87
Usage
Call the full DeepSeek lineup via OpenRouter Credits with deepseek/* model ids at models-page pass-through rates; high-value MoE models that rank among the most used on OpenRouter.
Models
Core models: DeepSeek V4 Pro 0813 flagship (~1M context, $0.435/$0.87), V4 Flash 0731 high throughput ($0.08/$0.18), V4 Pro ($1.168/$2.336), V4 Flash 0423 ($0.14/$0.28).
Highlights
Suited for coding, agents, and large-scale reasoning; 1M context at far lower cost than peer closed flagships for high-volume production.
xAI
Grok 4.x
Input
$1
Output
$2
Build entry · Grok 4.6 $2/$6
Usage
Call the full xAI Grok lineup via OpenRouter Credits with x-ai/* model ids at models-page pass-through rates—strong on real-time knowledge, coding, and STEM reasoning.
Models
Core models: Grok 4.6 / 4.5 flagship ($2/$6), Build 0.1 coding agent ($1/$2), Grok 4.3 reasoning ($1.25/$2.5), Grok 4.20 Multi-Agent ($1.25/$2.5).
Highlights
Suited for agent tool-calling, low-hallucination Q&A, and multi-agent workflows; one key switches among flagship, coding, and multi-agent tiers.
Qwen
Qwen3.8 / 3.7
Input
$0.03
Output
$0.13
3.7 Flash entry · 3.8 Max $2/$6
Usage
Call the full Qwen lineup via OpenRouter Credits with qwen/* model ids at models-page pass-through rates—strong bilingual Chinese/English, agent, and vision coverage from budget to flagship tiers.
Models
Core models: Qwen3.8 Max flagship ($2/$6), Qwen3.8 2.4T open MoE ($2/$6), Qwen3.7 Max ($1.475/$4.425), Plus ($0.32/$1.28), Flash ($0.03/$0.13).
Highlights
Suited for Chinese office work, coding, multimodal, and long-context production; one key switches among Flash/Plus/Max and open MoE tiers.
智谱 GLM
GLM 5.x
Input
$0.06
Output
$0.40
4.7 Flash entry · 5.2 $0.49/$1.54
Usage
Call the full Z.ai GLM lineup via OpenRouter Credits with z-ai/* model ids at models-page pass-through rates—strong for long-horizon agents, project-level software engineering, and Chinese workloads.
Models
Core models: GLM 5.2 flagship (~1M context, $0.49/$1.54), 5.1 long-horizon coding ($1.4/$4.4), 5V Turbo multimodal ($1.2/$4), 5 Turbo fast ($1.2/$4), 4.7 Flash budget ($0.06/$0.4).
Highlights
Suited for software-engineering agents, multi-step automation, and Chinese production; one key switches among Flash/Turbo/flagship and multimodal tiers.
MiniMax
M3 / M2.x
Input
$0.22
Output
$0.90
M2.5 entry · M3 $0.3/$1.2
Usage
Call the full MiniMax lineup via OpenRouter Credits with minimax/* model ids at models-page pass-through rates—focused on agent productivity, practical coding, and multimodal input.
Models
Core models: MiniMax M3 flagship multimodal (~1M context, $0.3/$1.2), M2.7 / M2.5 practical coding ($0.3/$1.2 · $0.22/$0.9), M2.1 light ($0.3/$1.2).
Highlights
Suited for long digital workflows, code generation, and image/video input; one key switches among M3 and M2.x tiers.
Kimi
K3 / K2.x
Input
$0.57
Output
$2.85
K2.5 entry · K3 $3/$15
Usage
Call the full Kimi lineup via OpenRouter Credits with moonshotai/* model ids at models-page pass-through rates—strong long-context and multimodal coding capabilities.
Models
Core models: Kimi K3 flagship (~1M context, $3/$15), K2.7 Code programming ($0.67/$3.4), K2.6 / K2.5 multimodal daily ($0.58/$2.44 · $0.57/$2.85).
Highlights
Suited for ultra-long docs, UI/code generation, and multi-agent orchestration; one key switches among K3 flagship and K2.x coding/daily tiers.
小米 MiMo
MiMo-V2.5
Input
$0.14
Output
$0.28
V2.5 entry · Pro $0.435/$0.87
Usage
Call the full Xiaomi MiMo lineup via OpenRouter Credits with xiaomi/* model ids at models-page pass-through rates—native omnimodal models at strong value and high OpenRouter usage.
Models
Core models: MiMo-V2.5-Pro flagship agent (~1.05M context, $0.435/$0.87), MiMo-V2.5 omnimodal daily ($0.14/$0.28).
Highlights
Suited for multimodal agents and high-throughput production; ~1M context at much lower cost than peer Pro tiers, one key switches Pro/standard.
Meta
Llama 4 / 3.x
Input
$0.10
Output
$0.30
Scout entry · Maverick $0.2/$0.70
Usage
Call the full Meta Llama lineup via OpenRouter Credits with meta-llama/* model ids at models-page pass-through rates—the most mature open-weight ecosystem for baselines and high throughput.
Models
Core models: Llama 4 Maverick flagship ($0.2/$0.70), Llama 4 Scout efficient ($0.1/$0.3), Llama 3.3 70B classic ($0.1/$0.32), Llama Guard 4 12B safety ($0.18/$0.18).
Highlights
Suited for open-weight baselines, high-throughput batching, and content-safety pipelines; one key switches Maverick/Scout/Guard.
Mistral
Large / Medium / Small
Input
$0.15
Output
$0.60
Small 4 entry · Medium 3.5 $1.5/$7.5
Usage
Call the full Mistral lineup via OpenRouter Credits with mistralai/* model ids at models-page pass-through rates—leading EU open and commercial models from general to edge.
Models
Core models: Mistral Medium 3.5 flagship ($1.5/$7.5), Large 3 general ($0.5/$1.5), Small 4 efficient ($0.15/$0.6), Ministral 14B / 8B edge ($0.2/$0.2 · $0.15/$0.15).
Highlights
Suited for general chat, EU-friendly use cases, and light/edge deployment baselines; one key switches Medium/Large/Small/Ministral.
Perplexity
Sonar search series
Input
$1
Output
$1
Sonar entry · Pro Search $3/$15
Usage
Call the full Perplexity Sonar lineup via OpenRouter Credits with perplexity/* model ids at models-page pass-through rates—focused on web search, deep research, and cited Q&A.
Models
Core models: Sonar Pro Search ($3/$15), Sonar Reasoning Pro ($2/$8), Sonar Pro ($3/$15), Sonar Deep Research ($2/$8), Sonar entry ($1/$1).
Highlights
Suited for research, support, and knowledge Q&A that need live retrieval and citations; one key switches Search/Reasoning/Deep Research tiers.
NVIDIA
Nemotron 3.x
Input
$0.05
Output
$0.20
Nano entry · Ultra $0.6/$3.6
Usage
Call the full NVIDIA Nemotron lineup via OpenRouter Credits with nvidia/* model ids at models-page pass-through rates—focused on agent orchestration, open reasoning, and enterprise high throughput.
Models
Core models: Nemotron 3.5 Lightning ($0.1/$0.25), Nemotron 3 Ultra ($0.6/$3.6), Nemotron 3 Super ($0.085/$0.4), Nemotron 3 Nano ($0.05/$0.2).
Highlights
Suited for agent orchestration, deep research, and enterprise high throughput; one key switches Lightning/Ultra/Super/Nano tiers.
腾讯
Hy3 / Hunyuan
Input
$0.06
Output
$0.21
Hy3 preview entry · Hy3 $0.132/$0.528
Usage
Call the full Tencent Hunyuan/Hy3 lineup via OpenRouter Credits with tencent/* model ids at models-page pass-through rates—strong for Chinese coding, finance, and document workloads, with high OpenRouter usage.
Models
Core models: Hy3 flagship ($0.132/$0.528), Hy3 preview entry ($0.063/$0.21), Hunyuan A13B Instruct ($0.14/$0.57).
Highlights
Suited for Chinese coding, finance analysis, and long-document work; one key switches Hy3 flagship/preview and Hunyuan tiers.

Openrouter Token Plan Price Comparison: Quota & Usage

Product type
Token Plan
Official product line
Credits Prepaid
5-hour limit
—
Weekly limit
—
Monthly limit
—

Openrouter Token Plan Price Comparison: Notes

  • Unused Credits may expire one year after purchase; refunds within 24 hours via Credits page (platform fees non-refundable; crypto never refundable).
  • Free model API defaults to 50 requests/day; rises to 1000/day after $10+ in purchased credits. Full catalog and unit prices at openrouter.ai/models.
  • Optional prompt/completion logging grants 1% usage discount; Privacy settings control provider training policy and data retention level.

Openrouter Token Plan Price Comparison: Tools & Integration

OpenRouter ChatREST APIOpenAI SDKClaude CodeCursor

Openrouter Token Plan Pricing, Top-up, Deals, Quota, Usage, Setup & Updates

View more updates

Page published：

Openrouter Token Plan Pricing, Top-up, Deals & FAQ

AI Token Plan

Global AI Token Comparison

AI Token Plan is not just a collection of vendor links. It places 41+ domestic and international AI platforms into one comparison framework, covering 395+ model entries plus common plan types such as Token Plans, Coding Plans, IDE Tools and LLM APIs.

When you need to compare AI subscription pricing, coding plan quota, API usage cost or official deal entry points, AI Token Plan brings official prices, plan tiers, usage rules, model capabilities and tool integrations into one place, reducing the need to check multiple vendor sites manually.

Data is continuously organized as vendor pricing pages, plan pages and product documentation change, making the homepage a pricing comparison entry point while detail pages explain whether each vendor plan fits individual developers, team purchasing or long-term API usage.

Prices come from each platform's official site and may change at any time; the official price prevails.

© 2026 AI Token Plan · All rights reserved · First published June 18, 2026 · 64 days running · Sitemap