Skip to main content
Everything you can do with ChatterMate

Every feature,
fully detailed.

From multi-model chat to a full cloud IDE, a content studio, enterprise controls, and a developer API — here's everything ChatterMate does.

10+
Frontier AI models
50+
Distinct features
5GB
Knowledge base per user
1-click
Deploy to production
01 · Multi-model Chat

Use every AI. In one thread.

Stop switching tabs. ChatterMate puts every frontier model in one conversation — swap at any point or ask them all simultaneously and compare responses side by side.

All frontier models

GPT-4o, Claude 4 Sonnet/Opus, Gemini 2.5 Pro/Flash, Grok 3 — all under one roof. New models added as soon as they launch.

  • Swap model mid-conversation
  • Per-model usage & cost tracking
  • Model capabilities panel

Ask All mode

Fire the same prompt to every model simultaneously. Get all responses in a tabbed view to compare reasoning, tone, and accuracy.

  • Tab through each response
  • Copy best answer instantly
  • Token cost shown per model

Smart routing

Auto-suggests the best model based on query complexity, context length, and your plan. Three-tier catalog cache keeps model list fresh.

  • Complexity score per query
  • Plan-aware access map
  • Override with one click

Shared context

Conversation context, uploaded files, and pinned artifacts travel with you as you switch between models. No re-uploading.

  • 14k–200k context windows
  • File attachments carry across models
  • Pinned system prompts

Conversation management

Search, pin, branch, and export conversations. Share a read-only link with a colleague or export the full thread as Markdown.

  • Full-text search across history
  • Branch from any message
  • Share link with expiry

Streaming & mobile

Real-time SSE streaming on every model. Flutter mobile client with OAuth PKCE, deep-link callbacks, and offline retry.

  • Animated streaming bubbles
  • QR preview via Expo Go
  • Auto-reconnect WebSocket
Multi-model chatAsk All
GPT-4o
Claude Sonnet
Gemini 2.5 Pro
Grok 3
U
GPT-4o

thinking…

Claude

thinking…

Smart routing·Complexity: high → suggested Claude 4 OpusSwitch →
GPT-4oClaude 4 SonnetClaude 4 OpusGemini 2.5 ProGemini FlashGrok 3
10+
Frontier models
1
Shared thread
<1s
First token
How it works
1
Ask oncesend to one model or all four at the same time
2
Compareanswers stream side-by-side with per-model cost
3
Keep goingswitch models mid-thread, context carries over
02 · Canvas & Artifacts

Write code, review it live.

Canvas is a persistent workspace that lives next to every chat. Ask AI to write or edit code, then review each change with a diff, accept or reject it, and preview the result — all without leaving the conversation.

CanvasPreviewDiff
auth.ts · reviewing…
async function verifyToken(tok) {
return jwt.verify(tok, secret)
const payload = jwt.verify(tok, secret)
if (!payload.sub) throw new Error()
return payload
}
AI suggested 3 changes
Validate sub claim
Return typed payload
Add expiry check
Revision history
v3 · 5m ago
v2 · 12m ago
⌘KCommand palette — refactor, explain, test…Mermaid · HTML · JSON
40+
Languages
Revisions kept
1-click
Accept / reject
How it works
1
Ask AI to editdescribe the change or hit ⌘K in the editor
2
Review the diffgreen/red inline, accept all or cherry-pick lines
3
Preview & rewindrender HTML/Mermaid live, roll back any version

Live code editor

Full-featured code editor with syntax highlighting for 40+ languages, bracket matching, and multi-cursor. Ctrl+K command palette for AI actions.

  • Syntax highlighting · 40+ langs
  • Ctrl+K AI command palette
  • Multi-cursor editing

AI diff with accept/reject

Every AI edit is presented as an inline diff — green additions, red removals. Accept all, reject all, or cherry-pick line by line.

  • Inline green/red diff
  • Accept / reject per chunk
  • Auto-commit to history

HTML & Mermaid preview

Renders HTML/CSS/JS output in a sandboxed iframe. Mermaid diagrams render live beside the code. Resize panes to match your workflow.

  • Sandboxed iframe preview
  • Live Mermaid diagram render
  • Resize & split panes

Revision history

Every save is versioned. Pin important checkpoints, compare any two versions side-by-side, and rewind to any state with one click.

  • Unlimited revision history
  • Pin & name checkpoints
  • Side-by-side version compare

Streaming AI edits

AI edits stream token-by-token directly into the editor. Watch the code appear in real time — no waiting for the full response.

  • Token-level streaming edits
  • Interrupt & redirect mid-stream
  • Cost estimate shown upfront

Templates & starters

Kick off new artifacts from built-in templates: React component, Express server, Prisma schema, Python script, and more.

  • 20+ built-in templates
  • Slash command to insert
  • Save custom templates
03 · RAG + Knowledge Base

Chat with your documents.

Drop any file or URL into ChatterMate's Knowledge Base and it becomes instantly queryable. Hybrid BM25 + pgvector retrieval with cross-encoder re-ranking means you get the most relevant passage — with inline citations — every time.

Universal file ingestion

PDF, DOCX, PPTX, XLSX, CSV, images, plain text, and live URLs. Async ingestion with real-time progress bars and chunk count.

  • PDF, DOCX, PPTX, XLSX, CSV
  • Image OCR extraction
  • Live URL scraping & indexing

Hybrid retrieval

Sliding-window chunking, BM25 keyword scoring, pgvector semantic embeddings, and RRF re-ranking combine for best-in-class precision.

  • BM25 + pgvector hybrid
  • RRF cross-encoder re-ranking
  • Configurable chunk overlap

Inline citations

Every answer that uses your documents includes clickable [N] citation markers. Click any to see the exact passage and page number used.

  • Clickable [N] citation markers
  • Passage + page number shown
  • Highlight source in original file

@mention any file

Type @filename in your prompt to direct a question at a specific document. Works across single files, folders, or the entire KB.

  • @mention to target a file
  • Cross-file synthesis
  • Folder-level @mention

Private vs. shared

Each file can be private (only you) or shared with your team workspace. Org admins can set default sharing policies.

  • Per-file visibility toggle
  • Team-shared collections
  • Org-level default policy

Storage & limits

100MB on Free, 5GB on Pro, 20GB on Team. Storage is calculated on compressed chunk size, not raw file size.

  • 100MB Free · 5GB Pro · 20GB Team
  • Compression on ingest
  • Usage meter in dashboard
Knowledge Base4 files · 128 chunks
PDFpolicy_v3.pdf2.4 MBready
DOCXarchitecture.docx840 KBready
XLSXmetrics_q3.xlsx1.1 MBindexing…
URLhttps://docs.acme.io14 pagesready
@policy_v3.pdf What are the data retention rules for EU customers?
retrieving passages…
Sources:[1] §4.2 p.8[2] §4.5 p.11[3] §7.1 p.19
7
File types
BM25+vec
Hybrid retrieval
[N]
Inline citations
How it works
1
Drop a file or URLPDF, DOCX, XLSX, images — indexed automatically
2
Ask in plain English@mention a doc to target it directly
3
Get cited answersevery claim links to the exact passage & page
04 · Image Generation

Generate images. Use them instantly.

ChatterMate integrates DALL·E 3, Gemini Imagen, and Grok image models under a single credit-based system. Generate, compare, and drop the result directly into Canvas or chat — no external tools needed.

Image Generationgenerating…
DALL·E 3Gemini ImagenGrok Image
Isometric illustration of a futuristic city at night, neon-lit streets, rain reflections, cinematic
Image credits149 / 200
3
Image models
4
Variants / prompt
200
Credits / mo
How it works
1
Describe itpick DALL·E, Imagen, or Grok and write a prompt
2
Compare variantsup to 4 renders side-by-side, keep the best
3
Use it instantlysend to Canvas or embed — unused variants refunded

Three generation models

DALL·E 3 for photorealistic renders, Gemini Imagen for stylised illustration, Grok image for creative concepts. Switch between them freely.

  • DALL·E 3, Gemini Imagen, Grok
  • Side-by-side comparison mode
  • Per-model style guide

Credit-based billing

Atomic reserve-and-refund ledger: credits are reserved before generation and only consumed on success. No partial charges on failure.

  • Atomic reserve & refund
  • Per-model credit cost shown upfront
  • 200 credits / mo on Pro

Canvas integration

Send any generated image to Canvas with one click. Use it as a design reference, annotate it with AI, or embed it in an HTML artifact.

  • Send to Canvas in 1 click
  • Annotate with AI overlay
  • Embed in HTML artifacts

Prompt history & retry

Every prompt and its generations are saved. Tweak the prompt and regenerate, or retry a failed generation without re-entering the prompt.

  • Full prompt history
  • Re-run with edited prompt
  • Auto-retry on API failure

Batch generation

Request up to 4 variants from the same prompt in one go. Compare side-by-side and keep only the best — unused variants are refunded.

  • Up to 4 variants per prompt
  • Unused variants refunded
  • Pick and discard in one click

Export & download

Download as PNG or WebP at full resolution. Every image is stored in your account library for 90 days.

  • PNG & WebP export
  • Full-resolution download
  • 90-day image library
05 · Teams & Enterprise

Workspaces with guardrails.

Invite your team, lock down which models they can use, and track every token spent — all from a single org admin panel. SOC-2-friendly audit logs, SSO via CleverCreator OAuth, and Stripe billing built in.

Org model allowlist

Admins can enable or disable any model for the entire org with one toggle. Users can only call models on the allowlist.

  • Per-model enable/disable
  • Policy inherited by all members
  • Instant enforcement (no re-login)

Usage dashboard

Per-member token breakdown with 7 / 14 / 30-day toggle and bar chart by model. Export CSV for finance.

  • Per-member breakdown
  • 7 / 14 / 30-day toggle
  • CSV export for finance

Roles & permissions

Owner, Admin, Member, Guest — each role controls which features, models, and data are accessible. Custom roles on Enterprise.

  • Owner / Admin / Member / Guest
  • Feature-level access gates
  • Custom roles on Enterprise

Audit & retention

Immutable token usage log, conversation exports, share-link expiry policies. Meets SOC-2 Type II audit requirements.

  • Immutable usage audit log
  • Conversation export (JSON/CSV)
  • Share-link expiry control

SSO + invites

CleverCreator OAuth 2.0 + PKCE. Email invite with configurable role. SAML on Enterprise. JIT provisioning supported.

  • CleverCreator OAuth 2.0 + PKCE
  • Email invite with role pre-set
  • SAML on Enterprise

Billing & plan control

Stripe-powered per-seat billing. Admins can cap monthly spend, set per-member token budgets, and receive over-limit alerts.

  • Per-seat Stripe billing
  • Monthly spend cap
  • Per-member token budget
RS
MK
LJ
+6
Acme · Engineering
9 members · Team plan · SSO on
Token usage · 14 daysΣ 1.42M
RS
Rohan S.
412k
MK
Maya K.
308k
LJ
Lily J.
241k
DT
Daniel T.
187k
EP
Esma P.
144k
Org model allowlist4 enabled · 3 restricted
GPT-4oClaude 4 SonnetGemini 2.5 ProClaude 4 OpusGPT-3.5Grok 2DALL·E 2
SSO
OAuth / SAML
SOC-2
Audit logs
Per-seat
Usage tracking
How it works
1
Invite your teamSSO via CleverCreator OAuth, roles per member
2
Set guardrailsallowlist which models each workspace can use
3
Track spendper-seat token usage and Stripe billing built in
06 · Developer API

One key. Every model.

Issue scoped API keys, call any model through a single REST + SSE endpoint, and embed ChatterMate into your own product. The same SSE protocol the web app uses — nothing proprietary, nothing surprising.

Developer API
from chattermate import ChatterMate
cm = ChatterMate(api_key="cm_live_…")
# Stream a response from Claude
async for chunk in cm.chat.stream(
model="claude-4-sonnet",
messages=[{"role":"user","content":"Hello"}],
):
print(chunk.delta, end="")
stablePOST /api/chat/streamSSE · all models
stablePOST /api/images/generatecredit-based
betaPOST /api/forge/sessionsspawn sandbox
betaGET /api/artifacts/{id}canvas export
1 key
Every model
SSE
Token streaming
2 SDKs
Python · Node
How it works
1
Issue a keyscoped cm_live_… key with expiry and pause/resume
2
Call one endpointPOST /api/chat/stream — same SSE the app uses
3
Ship itrate-limit headers, usage dashboard, CSV export

Scoped API keys

Generate cm_live_… keys with per-model scope, expiry dates, and pause/resume. Keys are SHA-256 hashed and Redis-cached (60s TTL).

  • Per-model scope
  • Expiry + pause/resume
  • SHA-256 hashed at rest

REST + SSE streaming

POST /api/chat/stream returns Server-Sent Events. Single endpoint for all models — one integration, every model.

  • POST /api/chat/stream
  • SSE token-level streaming
  • Same protocol as web app

Rate limiting

429 Retry-After responses with sliding-window per-action limits. Webhooks fire when you hit 80% of your monthly quota.

  • 429 + Retry-After
  • Sliding-window per-key
  • Webhook on quota warning

Images & Forge API

POST /api/images/generate for image generation (credit-deducted). POST /api/forge/sessions to spawn a Forge sandbox programmatically.

  • Image generation endpoint
  • Forge sandbox spawning (beta)
  • Artifact export endpoint (beta)

Usage dashboard

Per-key sparklines, total token spend, and per-model breakdown live in the developer dashboard. CSV export for billing reconciliation.

  • Per-key sparklines
  • Per-model token breakdown
  • CSV billing export

SDK & docs

Official Python and Node.js SDKs. Interactive API reference at /docs/api with request builder and live response viewer.

  • Python + Node.js SDK
  • Interactive /docs/api
  • Request builder & live test
07 · Forge IDE

Describe an app. AI ships it.

Forge is a full cloud development environment. Describe what you want in plain English — the AI agent scaffolds the app, writes every file, runs tests, and deploys to Cloudflare Pages or Vercel in one continuous workflow.

AI scaffolding agent

Describe your app in one sentence. The agent proposes an architecture plan, you approve, and it writes every file — components, API routes, DB schema, tests.

  • Natural language → full app
  • Architecture plan before writing
  • Multi-agent parallelism

Live in-browser preview

Auto-detected dev port renders inside a sandboxed iframe. Hot-reload on every save. No port-forwarding, no local setup.

  • Auto port detection
  • Hot-reload on save
  • Desktop / tablet / mobile frames

One-click deploy

Pre-deploy checkpoint, then push to Cloudflare Pages or your Docker VPS. SSL and custom domain wired automatically. Every deploy is reversible.

  • Cloudflare Pages · Vercel · Docker
  • Auto SSL + custom domain
  • One-click rollback

Sandboxed containers

Each project runs in an isolated container with Node 22, Python 3.13, and Postgres. Full filesystem access with no bleed between projects.

  • Isolated container per project
  • Node 22 · Python 3.13 · Postgres
  • Filesystem + package registry

Marketplace

Browse and install community-published starters — full-stack templates, UI kits, API wrappers. Publish your own with one command.

  • Full-stack community starters
  • UI kits + API wrapper templates
  • Publish your own starter

Self-testing agent

Headless Playwright tests run automatically after each build. The agent captures failures, writes fixes, and re-tests — bounded by iteration cap.

  • Auto Playwright test runs
  • Agent writes & applies fixes
  • Lighthouse + a11y scoring
Forge — build workflow
① Describe
"Build a CRM with Stripe & email notifications"
② Plan
12-step plan proposed · 7 / 12 steps complete
③ Build
Writing Dashboard.tsx · 42 files scaffolded
④ Test
Playwright · waiting…
⑤ Deploy
Cloudflare Pages · waiting…
Agent running · est. 90s remainingCost: 14k tok
Last deploy · 8m ago
https://acme-crm.pages.dev
Rollback ↺
08 · AI Writer

Write content that sounds like you.

AI Writer is a focused long-form writing environment — not a chatbot wrapper. Organise work in folders, use professional templates, write distraction-free in Focus Mode, and share finished documents with a public link.

AI WriterFocus Mode
The Complete Guide to Writing a PRD
Blog post·2,840 words·Published ↗

A product requirements document (PRD) is the contract between your team and the future. Done right, it eliminates ambiguity before a single line of code is written.

Start with the problem statement. Every stakeholder needs context before they can evaluate your solution — resist the urge to jump straight to features.

Next, define success metrics. If you can't measure the outcome, you can't confidently say the feature shipped successfully…

Templates
Blog post
PRD
SOP
Email
Landing page
Case study

Rich document editor

Markdown-powered editor with slash commands, inline AI completions, heading anchors, tables, and footnotes. Keyboard-first design.

  • Slash command menu
  • Inline AI completion (Tab)
  • Heading anchors & TOC

Templates library

Ready-made templates for blog posts, PRDs, SOPs, email sequences, landing page copy, case studies, and more. All fully editable.

  • 20+ professional templates
  • Save custom templates
  • Slash / to insert template

Folders & organisation

Nest documents in folders, star favourites, search across the full library by title or content, and sort by date or alphabetically.

  • Nested folders
  • Full-text content search
  • Star & pin favourites

Focus Mode

One keystroke hides the sidebar, nav, and all chrome. Just the document, a word count, and the AI. Ideal for long writing sessions.

  • Full-screen distraction-free mode
  • Word count + reading time
  • Dark, light, sepia themes

Share & export

Generate a public or password-protected share link. Export to Markdown, PDF, or clean HTML. Copy as rich text for pasting into Notion or Docs.

  • Public + password-protected link
  • Export: Markdown, PDF, HTML
  • Rich-text copy for Notion / Docs

AI rewrite & expand

Select any passage, then choose rewrite, shorten, expand, change tone, fix grammar, or translate — all via the inline AI toolbar.

  • Rewrite / shorten / expand
  • Tone: formal, casual, persuasive
  • Translate to 30+ languages
Ready to start

Every feature, free to try.

Create a free account and explore all eight feature areas — no credit card required. Upgrade when you're ready.