Guides & glossary
Practical guides for calling LLM APIs at the lowest price — and a plain-English glossary of the terms you meet along the way. Looking for the full price list? See the live pricing board.
Setting up a tool? Each guide under Set up your tool has the exact settings for that app. For Claude Code, see Claude Code on MartLLM; for Cursor, Goose, Raycast, RisuAI and more, see all integrations.
Articles
How to Make a Voice Over Video: Step-by-Step Guide
Learn how to make a voice over video: record clean narration, sync it with footage, add captions and export audio that sounds right on every device.
How to Choose and Integrate an Image Generation API
Compare image generation API capabilities, request fields, async results, model tests, error handling and billing, with steps for production integration.
Image to Animation: How to Animate a Still Image
This image to animation guide covers methods, step-by-step animation, frame checks and fixes for common motion failures in 2026.
How to Convert Link Into Video Without Losing Facts
Learn how to convert link into video in six steps: check the page, build a source-linked script, render clips, and review before publishing.
AI Generated Music Video: How It Works and Tools Used
Learn what an AI generated music video is, how the four-stage pipeline builds one, and which models and checks make it work.
Change Colors in Image Without Ruining Shading
Learn how to change colors in image elements while keeping shadows, texture and detail, with methods, steps and a colorblind check.
AI Video Generation API: Jobs, Models and Costs
Learn how an AI video generation API works, from async jobs and webhooks to model choice, output checks and true cost per usable clip.
List of AI Models 2026: 10 Options by Use Case
Browse this practical list of AI models by modality, capability, and use case, with 10 catalogs compared on pricing, access, and best-fit workloads.
Webhook vs API: Push, Polling and When to Use Each
Webhook vs API compared on direction, traffic, reliability and cost, with when to use each and how to combine them for async AI media jobs.
Text to Speech Download: Generate and Save Audio in 2026
Text to speech download guide: generate audio from text, save it as MP3 or WAV, check the file, and confirm you have the rights to use it.
Remove Background API: How It Works and What to Check
Learn how a remove background API returns transparent PNG masks, handles hard edges, and fits production image and video workflows.
How to Get Videos: Download, Stream, or Generate
Learn how to get videos legally: download files you have rights to, stream from platforms, or generate AI clips, with costs, rights checks and fixes.
Text to Speech Models: Capabilities and Selection Guide
Explore text to speech models by input, output, voice cloning, quality metrics and consent rules, with tables and a selection checklist for 2026.
How Much Does Modeling Cost? AI Inference Prices Explained
How Much Does Modeling Cost in 2026? Current AI per-token and per-job prices, a monthly bill formula, and the hidden costs that change it.
Singing Voice Synthesis: How It Works and Where It Fails
Learn how singing voice synthesis turns lyrics and notes into sung audio, what training data and vocoders control, and where it still falls short.
Screenshot to Video: How to Animate a Still Image Safely
Learn screenshot to video the right way: choose the output, prep the image, redact sensitive text, and check motion before you publish.
AI Pricing Models Explained: Tokens, Tasks and Costs
Compare AI pricing models: per-token, per-task, subscription and credit billing, plus how to estimate what your inference will really cost.
Photo Editing API Guide: Features, Formats and Workflows
A practical guide to photo editing API features, formats, async jobs, retries, and provenance so you can integrate with confidence.
How to Make AI Music Video Synced to a Track
How to make AI music video clips that stay on the beat, from the shot list and short generations to lip sync, rights, and the final cut.
Create a 3D Image: Methods, Steps and Fixes for 2026
Learn how to create a 3D image with AI models, photogrammetry and 3D tools, plus fixes for common problems and a usability checklist.
Change Landscape Photo to Portrait: Crop, Pad or Extend
Learn how to change landscape photo to portrait without losing the subject, when to crop, pad or extend, and how to check AI edits.
10 Text to Video Prompts for Stronger AI Clips
10 practical text to video prompts with examples, tips, and trade-offs to help you generate cleaner, more consistent AI video results.
Set up your tool
Aider with a cheaper OpenAI-compatible API (MartLLM)
Run Aider on MartLLM with OPENAI_API_BASE and --model openai/<id>, plus the model settings file that turns on diff edits and the repo map for models Aider does not know yet.
Chatbox with MartLLM: one-click import or 1-minute manual setup
Add MartLLM to Chatbox (desktop and mobile) as an OpenAI API Compatible provider, with a one-click chatbox:// import link. Then use GPT, Claude, Gemini, DeepSeek and more at market price.
Cline with a cheaper API: MartLLM as an OpenAI Compatible provider
Set Cline’s API Provider to OpenAI Compatible with base URL https://api.martllm.com/v1 and your MartLLM key. Exact field names, which models handle Cline’s tool calls, and fixes for an empty model list.
Codex CLI with MartLLM: run OpenAI’s coding agent on any model
Point OpenAI’s Codex CLI at MartLLM with a custom model provider in ~/.codex/config.toml (wire_api = "responses"). Runs DeepSeek, GLM, Kimi, Qwen Coder, Claude or GPT at market price, with tool calls and file edits.
Continue (VS Code / JetBrains) with MartLLM: config.yaml that works
Add MartLLM models to Continue’s config.yaml with provider openai and apiBase https://api.martllm.com/v1, plus the two settings that stop 404s on autocomplete and GPT-5 models.
Janitor AI proxy setup with MartLLM (cheap uncensored models)
Use MartLLM as your Janitor AI proxy: the exact Proxy URL, model names that work, and the save-and-refresh steps from Janitor’s own help center. Uncensored roleplay models at market price.
Kilo Code with MartLLM: add a custom provider in 1 minute
Point Kilo Code (VS Code, JetBrains and the CLI) at MartLLM: Settings → Providers → Custom provider, or kilo.jsonc. Fetch the model list, set context limits, and run top coding models at market price.
LibreChat custom endpoint for MartLLM (librechat.yaml)
Add MartLLM to LibreChat as a custom endpoint in librechat.yaml: the exact config, the .env line, Docker mount, and live model picks. GPT, Claude, Gemini and open models from one key.
n8n with MartLLM: cheaper AI Agent models and video generation
Use MartLLM in n8n: an OpenAI credential with Base URL https://api.martllm.com/v1 for the AI Agent’s OpenAI Chat Model (Responses API off), and an HTTP Request workflow that makes videos with polling.
Open WebUI with MartLLM: every model from one connection
Add MartLLM to Open WebUI as an OpenAI API connection (or with OPENAI_API_BASE_URL in Docker) and get GPT, Claude, Gemini, DeepSeek and more in the model picker at market price.
OpenCode with MartLLM: a cheaper provider in opencode.json
Add MartLLM to OpenCode as a custom provider (@ai-sdk/openai-compatible) in opencode.json, store the key with /connect, and run Claude, GLM, Kimi, DeepSeek or Qwen Coder at market price.
SillyTavern with a cheaper API: connect MartLLM in one minute
Connect SillyTavern to MartLLM through Chat Completion → Custom (OpenAI-compatible). Exact field names, uncensored roleplay models with live prices, and fixes for "no models" and blank replies.
Use the OpenAI Python SDK and LangChain with a cheaper API (MartLLM)
Change two arguments, base_url and api_key, and the official openai Python SDK or LangChain ChatOpenAI runs on MartLLM: GPT, Claude, DeepSeek, GLM, Kimi and more from one key at market price.
Use the Vercel AI SDK with MartLLM (cheaper models, same code)
Point the Vercel AI SDK at MartLLM with @ai-sdk/openai-compatible: one provider file, then generateText, streamText, useChat and tool calls work on every chat model at market price.
Cost & pricing
The cheapest Claude API (and how to switch in one line)
A practical guide to calling Claude (Opus, Sonnet, Haiku) for the lowest price: route every request to the cheapest healthy seller through one OpenAI- or Anthropic-compatible key. Live Claude prices inside.
The cheapest GPT API (route it, or switch models in one line)
How to call GPT models for the lowest price — routed to the cheapest healthy seller — or drop to far cheaper open-weight models with a one-line change. OpenAI-compatible, one key, no fee added. Live GPT prices inside.
Compatibility & integration
One API key for all LLMs: stop juggling provider accounts
Replace separate OpenAI, Anthropic, Google and DeepSeek keys with one OpenAI- and Anthropic-compatible key that reaches every major LLM — routed to the cheapest healthy seller, no fee added. Live prices inside.
What is an OpenAI-compatible API endpoint?
An OpenAI-compatible API endpoint accepts the same requests and returns the same responses as the OpenAI API, so the OpenAI SDK works against it by changing only the base URL. Here is what that means in practice.
How MartLLM works
How MartLLM catches a seller listing the wrong model
Announcement: how MartLLM checks that a seller’s offer really serves the model it lists. The model is matched at listing, tested before any buyer, reviewed by a person, and tested again on live requests. A seller caught faking is banned and its buyers are refunded.
Smart routing: the cheapest provider that is fast enough for your prompt
How MartLLM picks a provider for every request: it measures each provider’s time to first token at your prompt size, prices your exact request on each one, and sends it to the cheapest provider that is fast enough. Long prompts also get a second provider when the first is slow to start.
How MartLLM (formerly UberLLM) routing picks the cheapest healthy provider
A clear explanation of how MartLLM (formerly UberLLM) routes each request: it ranks the sellers of your model by live price, skips unhealthy ones, sends the request to the cheapest healthy seller, and retries on the next if it fails before output.
Sell your credits
Sell unused AI credits: list a provider key in two minutes
How to list unused AI provider credit on MartLLM from the terminal: one command checks your key with the provider, shows the balance where the provider has an API for it, and lists it only after you say yes.
Sell unused AkashML credit: check the balance and list it
How to list unused AkashML credit on MartLLM with one command: the key is checked with AkashML’s model list, you read the balance in your AkashML account, and what AkashML’s terms say about reselling.
Sell unused Alibaba DashScope — China credit: check the balance and list it
How to list unused Alibaba DashScope — China credit on MartLLM with one command: the key is checked with Alibaba DashScope — China’s model list, you read the balance in your Alibaba DashScope — China account, and what Alibaba DashScope — China’s terms say about reselling.
Sell unused Amazon Bedrock credit: check the balance and list it
How to list unused Amazon Bedrock credit on MartLLM with one command: the key is checked with Amazon Bedrock’s model list, you read the balance in your Amazon Bedrock account, and what Amazon Bedrock’s terms say about reselling.
Sell unused Bankr credit: check the balance and list it
How to list unused Bankr credit on MartLLM with one command: the key is checked with Bankr’s model list, you read the balance in your Bankr account, and what Bankr’s terms say about reselling.
Sell unused Capminal Gateway credit: check the balance and list it
How to list unused Capminal Gateway credit on MartLLM with one command: the key is checked with Capminal Gateway’s model list, you read the balance in your Capminal Gateway account, and what Capminal Gateway’s terms say about reselling.
Sell unused Concentrate.ai credit: check the balance and list it
How to list unused Concentrate.ai credit on MartLLM with one command: the key is checked with Concentrate.ai’s model list, you read the balance in your Concentrate.ai account, and what Concentrate.ai’s terms say about reselling.
Sell unused DeepSeek credit: check the balance and list it
How to list an unused DeepSeek balance on MartLLM with one command: the balance check reads paid-in and granted credit from DeepSeek’s own API, and what DeepSeek’s platform terms say about sharing a key.
Sell unused Engy.ai credit: check the balance and list it
How to list unused Engy.ai credit on MartLLM with one command: the key is checked with Engy.ai’s model list, you read the balance in your Engy.ai account, and what Engy.ai’s terms say about reselling.
Sell unused Fireworks AI credit: check the balance and list it
How to list unused Fireworks AI credit on MartLLM with one command: the key is checked with Fireworks AI’s model list, you read the balance in your Fireworks AI account, and what Fireworks AI’s terms say about reselling.
Sell unused Groq credit: check the balance and list it
How to list unused Groq credit on MartLLM with one command: the key is checked with Groq’s model list, you read the balance in your Groq account, and what Groq’s terms say about reselling.
Sell unused Meta Model API credit: check the balance and list it
How to list unused Meta Model API credit on MartLLM with one command: the key is checked with Meta Model API’s model list, you read the balance in your Meta Model API account, and what Meta Model API’s terms say about reselling.
Sell unused Mistral AI credit: check the balance and list it
How to list unused Mistral AI credit on MartLLM with one command: the key is checked with Mistral AI’s model list, you read the balance in your Mistral AI account, and what Mistral AI’s terms say about reselling.
Sell unused Moonshot AI / Kimi credit: check the balance and list it
How to list an unused Kimi (Moonshot AI) balance on MartLLM with one command: the check reads cash and voucher balance from the Kimi platform API, which platform’s keys work, and what the Kimi terms say about selling keys.
Sell unused Mordiem credit: check the balance and list it
How to list unused Mordiem credit on MartLLM with one command: the key is checked with Mordiem’s model list, you read the balance in your Mordiem account, and what Mordiem’s terms say about reselling.
Sell unused Morpheus credit: check the balance and list it
How to list unused Morpheus API credit on MartLLM with one command: the check reads paid credit and today’s staking credit from the Morpheus billing API, and where to read the Morpheus terms before you list.
Sell unused Ollama Cloud credit: check the balance and list it
How to list unused Ollama Cloud credit on MartLLM with one command: the check reads purchased credit and what is left in your plan from Ollama’s balance API, and what to ask Ollama before you list.
Sell unused OpenRouter credit: check the balance and list it
How to list unused OpenRouter credit on MartLLM with one command: what the key check reads from OpenRouter, how to cap what a key can spend, and what OpenRouter’s terms say about reselling.
Sell unused Together AI credit: check the balance and list it
How to list unused Together AI credit on MartLLM with one command: Together has no balance API for a key, so the check uses the model list, how to limit what an offer can spend, and what Together’s terms say about reselling.
Sell unused TypeSafe credit: check the balance and list it
How to list unused TypeSafe credit on MartLLM with one command: the key is checked with TypeSafe’s model list, you read the balance in your TypeSafe account, and what TypeSafe’s terms say about reselling.
Sell unused Uncensored AI credit: check the balance and list it
How to list unused Uncensored AI credit on MartLLM with one command: the key is checked with Uncensored AI’s model list, you read the balance in your Uncensored AI account, and what Uncensored AI’s terms say about reselling.
Sell unused Venice AI credit: check the balance and list it
How to list unused Venice AI credit on MartLLM with one command: the check reads your USD, bundled and Diem balances from Venice’s own API, and what Venice’s terms say about selling access.
Sell unused Z.ai Coding credit: check the balance and list it
How to list unused Z.ai Coding credit on MartLLM with one command: the key is checked with Z.ai Coding’s model list, you read the balance in your Z.ai Coding account, and what Z.ai Coding’s terms say about reselling.
Sell unused Z.ai credit: check the balance and list it
How to list unused Z.ai credit on MartLLM with one command: the key is checked with Z.ai’s model list, you read the balance in your Z.ai account, and what Z.ai’s terms say about reselling.