Typeshoot
Images, video, voiceovers, music and editing inside Claude and ChatGPT, priced in credits. Generate and edit images and video, make voiceovers, music and sound effects, add captions and remove backgrounds, with a library and brand kits.
託管 MCP 伺服器
npx add-mcp 'https://typeshoot.com/mcp'可安裝到 Claude Code、Codex、Cursor 等客戶端
文件
Endpoint
https://typeshoot.com/mcp
JSON-RPC 2.0 over Streamable HTTP. Each request gets a single JSON response; there is no SSE stream yet, which the spec allows. Results that render media include an MCP Apps card, so hosts that support it show the image, clip or audio inline.
Authentication
- OAuth 2.1 for assistants: dynamic client registration, PKCE, refresh and revoke. Clients discover it from the
WWW-Authenticateheader and/.well-knownmetadata, so pasting the URL is enough. - API keys for scripts, CI and headless agents: create one in your dashboard and send it as
Authorization: Bearer <key>. Keys act as you and spend your credits; revoke them there.
Try it
curl -s https://typeshoot.com/mcp \
-H "Authorization: Bearer $TYPESHOOT_API_KEY" \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'
Billing
Every call that spends credits reserves its estimate first, settles at the real cost and refunds automatically if the job fails on our side. Call estimate_generation to price a job without running it, and get_credits_balance to check what’s left. Prices are on the pricing page and per model on /models.
Tools served today (28)
Create
| Tool | What it does | Credits |
|---|---|---|
generate_image | Create a new image from a text prompt, or edit one by passing image URLs. Works with Nano Banana, GPT Image, FLUX, Seedream, Recraft and more. | From 6 credits |
generate_video | Produce video from text, from a single start frame, or from first and last frames. Draft at 480p first, then render the one you like. | From 5 credits a second |
generate_speech | Turn any text into studio-quality narration with a voice, a language and a delivery. Use a stock voice, a designed one or a saved Character’s. | 20 credits per 1,000 characters |
generate_music | Background music, jingles or full songs. Lyria for quick beds, ElevenLabs Music for an exact length, MiniMax or Mureka to sing your lyrics. | 20 credits a track; 120 a minute for an exact length |
generate_sound_effect | Sound effects, Foley, ambience and UI sounds from a description, 0.5 to 30 seconds, with seamless loops for ambience. | About 0.4 credits a second |
recommend_video_model | Reads your brief (length, shape, sound, speech, start image, style, budget) and returns the best three models with exact credits and trade-offs. | Free |
Edit & transform
| Tool | What it does | Credits |
|---|---|---|
upscale_image | Enlarge an image up to 4× with recovered detail, ready for print or a hero banner. | 2 or 80 credits |
remove_background | Cut the subject out of an image or a video and return it on transparency, ready for a new scene. | 2 credits an image |
trim_video | Cut a clip to a start and end time, or to several segments in order. | 2 credits |
reframe_video | Change the shape, for example 16:9 to 9:16 for TikTok and Reels, by cropping to fill the frame or fitting it with bars. | 2 credits |
add_captions | Burn word-by-word captions into a video, TikTok style. Uses the timings from a voiceover, or transcribes the speech itself. | 6 credits a minute, plus ~2 to transcribe |
Combine
| Tool | What it does | Credits |
|---|---|---|
merge_media | Concatenate clips, lay audio on video, or mix music under a voiceover with ducking. | 2+ credits |
Characters
| Tool | What it does | Credits |
|---|---|---|
swap_video_characters | Replace a person in an existing video with a saved Character, keeping the performance and camera move. | From 10 credits a second |
Your library
| Tool | What it does | Credits |
|---|---|---|
search_assets | Find anything you’ve made or uploaded by words, type, tag or date. | Free |
get_upload_url | Get a private upload link for files too big to send through the chat. | Free |
confirm_upload | Tell Typeshoot the upload finished, so the file appears in your library. | Free |
analyze_video | Breaks a video into shots with timestamps, shows the key frames and transcribes the speech, so the assistant can explain the hook, structure and pacing before you recreate or swap it. | About 2 credits a minute |
import_media | Bring in media from a public video or social link, ready to edit or transcribe. | 1 credit |
Brand themes
| Tool | What it does | Credits |
|---|---|---|
update_theme | Create or change a theme from uploads, your website or a description. | Free |
get_theme | Read a theme, so your assistant knows the rules before it writes a prompt. | Free |
list_themes | See every theme, handy when you work for more than one brand. | Free |
delete_theme | Remove a theme you no longer need. | Free |
Discovery & account
| Tool | What it does | Credits |
|---|---|---|
list_models | Browse the models for a task, with what each is good at and its price. | Free |
get_model_params | Read one model’s inputs, limits and defaults before using it. | Free |
estimate_generation | Price a generation before it runs. Anything over 100 credits asks first. | Free |
check_job | Get a job’s status, and its permanent file link once it’s done. | Free |
get_credits_balance | See your balance and what you’ve spent recently. | Free |
get_credits_link | Get a checkout link for a credit pack. Credits never expire. | Free |
Try it in your own chat
Typeshoot adds image, video, voice and editing tools to Claude, ChatGPT and other assistants. 100 free credits to start, no card needed.