Models
Use the deepseek-flash model string to access DeepSeek V4.1 Flash for text, coding, vision, and agent workflows. One GPTProto API key and balance also give you access to 200+ models for evaluation, routing, and fallback design.
Access Z.ai’s first natively multimodal GLM-5 model through GPTProto. Send text, images, videos, or files, retain up to 1M tokens of context, and use one API key and shared balance across 200+ supported models.
Access GPT-6 Luna on GPTProto for high-volume chat, classification and structured extraction. Run focused tasks with adjustable reasoning and a 1.05M-token context window. Check the live rate and test the model using your GPTProto API key.
Access Z.ai's GLM-5.3 API through GPTProto for coding, defensive code review, and long-horizon agent workflows. Get 1M-token context, up to 128K-token output, OpenAI-compatible access, and 10% lower token pricing with one key shared across 200+ models.
Access Tencent's 770B-parameter Hy4-preview API through GPTProto for long-context coding, document analysis, and agent workflows. Use one API key and one shared balance to test it alongside 200+ AI models.
Access Anthropic's Claude Opus 5.5 for coding, long-running agents, and document analysis through GPTProto. The model accepts text and images, provides a 1M-token context window and up to 128K output tokens, and uses adaptive thinking. Check the live pricing panel for GPTProto's current rates.
Access Moonshot AI's Kimi K3.1 Pro through GPTProto for repository-scale coding, visual-to-code tasks, financial research, and long-running agents. Use one GPTProto key and balance to test it alongside 200+ text, image, video, and audio models.
Access GPT 6.1 Sol through GPTProto for coding, document analysis, and multi-step agent work. OpenAI positions the model near Astra for demanding tasks at lower standard token rates. Use one GPTProto key and balance to compare 200+ models across providers.
Kimi K3 is Moonshot AI's 2.8T-parameter multimodal reasoning model for long-horizon coding and knowledge work. GPTProto provides Kimi K3 API access at $2.70 input and $13.50 output per 1M tokens, 10% below Moonshot's current list prices.
GLM-5.2 is Z.ai's (formerly Zhipu AI) open-weight, 753B-parameter Mixture-of-Experts model with a lossless 1M-token context window, trained for long-horizon agentic coding and repository-wide refactoring. Call the GLM-5.2 API on GPTProto at $1.26 / $3.96 per 1M tokens — 10% under Z.ai's list rate, with one key and one …
Access Anthropic's Claude Sonnet 5.5 through GPTProto for coding, agent workflows, document analysis, and visual understanding. Compare its live token rate with other models using one API key and shared balance.
Access Google's Gemini 4 Argon through GPTProto for text reasoning, software engineering, long-horizon agents, and structured analysis. Call gemini-4-argon with one GPTProto key and one shared balance across 200+ models, with API rates 40% below Google's direct pricing.
Access the current DeepSeek-V4-Flash-0731 model through GPTProto for coding, tool-driven agents, long-context review, and high-volume text workloads. Use one OpenAI-compatible API key and a shared balance across DeepSeek V4 Pro and 200+ other models.
Access OpenAI's GPT-6.1 Terra through GPTProto for visual reasoning, computer use, coding, and natural writing. Get a 1.05M-token context window, adjustable reasoning, and 20% lower token rates with one key shared across 200+ models.
Access Google's Gemini 3.8 API through GPTProto for long-horizon coding, autonomous agents, and multimodal analysis. Use a 1M-token context window, 64K maximum output, three thinking levels, and 40% lower token pricing with one balance shared across 200+ models.
Access Anthropic's Claude Haiku 5.5 through GPTProto for high-volume classification, extraction, routing, support, and subagent work. Test it with one key and a shared balance across available models.
gpt-5.6-luna is the fastest and cheapest tier of OpenAI's GPT-5.6 family, built for cost-sensitive, high-volume workloads: classification, extraction, summarization, and user-facing chat. The gpt 5.6 luna api on GPTProto costs $0.80/$4.80 per 1M tokens — 20% under OpenAI's $1/$6 list price.
DeepSeek 4 Pro API delivers flagship-level reasoning with a 1M context window. Optimized for agentic coding and STEM logic, it offers elite performance at 1/8th the cost of competitors. Access the deepseek 4 pro api via GPTProto.com today.
MiniMax M3 is a frontier Mixture-of-Experts model featuring a 1M token context window and native multimodal support. Built for high-fidelity reasoning, MiniMax M3 excels in coding, bilingual tasks, and long-document analysis.
Access Grok 4.7 through GPTProto for coding, knowledge work, image understanding, and long-running agents. Use a 500K-token context window, configurable reasoning, and one shared API key and balance across 200+ AI models.
No model matches these filters.
Scenario: Programming
| Output | Stability | Action | ||||
|---|---|---|---|---|---|---|
 DeepSeek Flash20% off official | 1.05M | |||||
 GLM 5.3 Flash15% off official | 1.05M | |||||
 GPT 6 Luna20% off official | 1.05M | |||||
GLM 5.315% off official | 1.05M | |||||
 Hy4 Preview5% off official | 1.05M | |||||
 Claude Opus 5.510% off official | 1M | |||||
| 1.05M | ||||||
GPT 6.1 Sol20% off official | 1.05M | |||||
 Kimi K315% off official | 1.05M | |||||
GLM 5.215% off official | 1.05M | |||||
Claude Sonnet 5.510% off official | 1M | |||||
 Gemini 4 Argon40% off official | 1.05M | |||||
DeepSeek v4 Flash20% off official | 1.05M | |||||
GPT 6.1 Terra20% off official | 1.05M | |||||
Gemini 3.8 Flash40% off official | 1.05M | |||||
Claude Haiku 5.510% off official | 1M | |||||
GPT 5.6 Luna20% off official | 1.05M | |||||
DeepSeek v4 Pro20% off official | 1.05M | |||||
 MiniMax M320% off official | 1.05M | |||||
 Grok 4.740% off official | 500K |
New model, new key.One key, all models.Priced below official.
Every model's modalities, price and discount in one list — call any of them on the key you already use.
Gemini 3.8 Flash Tts40% offAccess Google’s flagship creative text-to-speech model through GPTProto for expressive narration, multi-speaker dialogue, voice design, and voice replication. Generate audio in more than 130 languages with one API key and pricing 40% below Google’s current paid-tier rate.
$10.80 out
$18.00 out · save $7.20
GPT 6.1 Terra20% offAccess OpenAI's GPT-6.1 Terra through GPTProto for visual reasoning, computer use, coding, and natural writing. Get a 1.05M-token context window, adjustable reasoning, and 20% lower token rates with one key shared across 200+ models.
$4.00 out
$5.00 out · save $1.00
Kimi K3.1 Pro HighspeedAccess Moonshot AI's faster K3.1 Pro serving tier through GPTProto for interactive coding, responsive research, and agent loops where each step waits for the previous result. Use one GPTProto key and balance to route between Highspeed, the lower-cost standard tier, and other supported models.
$20.00 out
$20.00 out
Kimi K3.1 ProAccess Moonshot AI's Kimi K3.1 Pro through GPTProto for repository-scale coding, visual-to-code tasks, financial research, and long-running agents. Use one GPTProto key and balance to test it alongside 200+ text, image, video, and audio models.
$10.00 out
$10.00 out
Kling 4.0Kling 4.0 is Kuaishou's native multimodal video model for 3–30-second generation, 10-keyframe direction, multimodal references, synchronized stereo audio, and output up to 4K. Access it through GPTProto with one API key and shared balance.
What might your work cost?
See what to add to your balance.
Tell us what you're working on and how often you use the API. We'll suggest a starting balance, and show an example of what it could cover.
Video and agent workloads burn credits faster — the recommendation adjusts automatically as you switch scenario or usage level.
Pick the vendor you know.See what they make, and what you save.
All 22 vendors we carry. Open a tile for that vendor's models.
OpenAI57 modelsThe Brainiac of the Bunch，Think of GPT as the friend who always has the right words—whether you’re writing, coding, or brainstorming. The GPT API delivers powerful text generation and understanding. It’s like hiring a genius assistant, without the coffee breaks.30% offGoogle33 modelsThe Dual-Thinking Dynamo，From Google’s genius labs comes Gemini—a model built for multi-modal mastery. With the Gemini API, you’re not just thinking big—you’re thinking in pixels, prose, and precision.60% off
Claude24 modelsThe Polite Powerhouse，Claude is like your super-smart, ultra-considerate buddy who never interrupts but always impresses. With the Claude API, you get deep, thoughtful responses and refined conversation. Formal or casual, Claude’s got range.10% off
Bytedance23 modelsBytedance is your go-to teammate for sharp, fast, and context-aware dialogue. Whether it's customer service or workplace Q&A, Doubao adapts like a pro. Think of it as the colleague who always gets it—and never misses a beat.40% off
Grok21 modelsThe Snarky Genius from X，Elon’s got jokes—and Grok’s got brains. The Grok API delivers sharp, contextual AI responses with a hint of attitude. It’s like chatting with the internet’s favorite rogue AI—minus the chaos.40% offKling18 modelsHollywood’s Favorite AI Filmmaker，Turn scripts into scenes in seconds. The Kling API helps creators produce cinematic videos with jaw-dropping realism. Perfect for marketers, storytellers, and future Spielbergs.20% off
MiniMax16 modelsMiniMax develops general artificial intelligence technology, offering text-to-video models, AI chatbots, and music generation through consumer applications.40% off
Qwen14 modelsQwen (short for "Tongyi Qianwen") is a state-of-the-art large language model (LLM) family developed by Alibaba Cloud. It is designed to handle a wide range of natural language understanding and generation tasks across multiple languages, especially Chinese and English.30% off
Vidu8 modelsVidu is a generative media platform operated by ShengShu AI HK Limited, with a video-first product line spanning text-, image-, reference-, and frame-guided creation. On GPTProto, the Vidu models collection covers supported video routes plus Q2 image generation and editing workflows. Q labels indicate model generations, while Pro, Turbo, and Pro Fast identify different quality, speed, or cost positions rather than separate providers. GPTProto carries only the AI models by Vidu that are currently supported on the platform, not Vidu’s entire official catalog. The directory updates as GPTProto adds, changes, or removes supported routes.30% off
DeepSeek7 modelsDeepseek is revolutionizing the way people search and interact with data. With its powerful AI-driven search engine, Deepseek helps users instantly find the most relevant information from vast datasets, making research, decision-making, and insights faster than ever. Perfect for analysts, researchers, and anyone looking to uncover insights with speed and accuracy.40% off
Z-AI6 modelsZ.ai develops the GLM family of language and multimodal AI models, with a strong focus on coding, reasoning, tool use, and agentic workflows. Z-AI is the provider label used in GPTProto, while GLM identifies the model family shown in the catalog. This Z-AI models collection brings the supported GPTProto routes, live prices, Model IDs, and task fit into one place for practical comparison. Use the cards above to review current availability rather than treating this page as a list of every model released by Z.ai. The catalog is updated as GPTProto adds, changes, or removes supported routes.15% off
GPTProto5 modelsStable and Affordable API Access to All Top AI Models. Experience a powerful API platform with unmatched stability, competitive pricing, and unified access to every model — all with a single key.
Ideogram5 modelsThe Typographer’s Dream Tool，Need words that wow? The Ideogram API transforms text into visual art—logos, stylized designs, and more. Say it loud, say it with flair.20% offMoonshotAI5 modelsMoonshot AI develops Kimi, a series of open-weight multimodal language models for coding, reasoning, knowledge work, tool use, and visual understanding. This MoonshotAI models collection brings together the Kimi APIs currently supported on GPTProto, allowing developers to compare live Model IDs, context limits, capabilities, and token pricing in one place. The catalog is updated as new AI models by MoonshotAI become available, while the same GPTProto account and shared balance can also be used across other supported providers.15% off
Higgsfield3 modelsHiggsfield is an advanced AI research and deployment company focused on generative video and image technologies. By leveraging state-of-the-art diffusion models, Higgsfield empowers creators and developers to generate high-quality visual content from simple text or image prompts.30% off
Flux2 modelsFLUX is a cutting-edge AI model designed to deliver intelligent, quick, and adaptive responses with a flair for creativity and personality. Think of it as your go-to digital companion for problem-solving, brainstorming, and making things more engaging with a touch of wit.20% off
Meta2 modelsMeta develops AI models for language, reasoning, and multimodal understanding. Its Muse models support work such as coding assistance, analysis, and applications that use tools. Muse Wonder and Muse Spark have different workload priorities, helping teams balance demanding tasks against the cost of repeated requests. Meta also develops the Llama family; availability on GPTProto depends on the specific model. This Meta models collection focuses on models supported through GPTProto, with model pages for pricing and integration details. The catalog updates as supported availability changes, so check the selected model’s status and request requirements before deployment.20% off
Hunyuan1 modelsHunyuan, also branded Tencent Hy, is Tencent's model family. Tencent develops separate language, image, video, and 3D systems under related names, so a familiar Hunyuan label does not mean every model accepts the same inputs or uses the same API. This directory shows only the Hunyuan models currently supported on GPTProto, with each model's live capabilities, price, and route details on its card. The language-model route can assist with coding, analysis across documents, and research workflows, while the surrounding application remains responsible for tools and verification. Check the model card before choosing a request format; the catalog changes as supported routes change.5% off
Midjourney1 modelsThe Digital Dream Painter，Turn your wildest visuals into stunning art. The Midjourney API brings imagination to life through AI-generated images that pop off the screen. Ideal for creatives, designers, and digital wizards.40% off
Suno1 modelsFrom Text to Tracks，Make music with your mind. The Suno API turns words into beats, melodies, and full-blown soundscapes. From jingles to bangers, you’re now the producer.20% off
Tripo3d1 modelstripo3d
TypesafeAI1 modelsTypeSafe AI develops System One models for decisions embedded in software. Jev is its decision-model family, designed to evaluate supplied information against questions defined by an application. A support platform can classify an incoming message, assess urgency, or check whether it meets a condition before choosing the next step. The application defines the options or scoring rubric and decides how to use the result. This TypesafeAI models collection covers models supported by GPTProto and updates as availability changes. Open a model page to review its supported decision interface, current pricing, and deployment details.20% offYour Usage.Your Potential Savings.
Choose a model and enter your monthly usage to compare estimated costs on GPTProto, the official provider, and OpenRouter.
Pay Per Token.
No Subscription. No Minimum.
You only pay for what you use. Top up once, spend it on any of 200+ models.
Get started. Great for testing the endpoint and trying new models.
The sweet spot for individual developers and small projects shipping to production.
Built for teams running production workloads. Bonus credits stack on already-discounted model pricing.