Technical Setup Guide

The Higgsfield MCP Connector

Thirty-plus image and video models behind one connection. Setup, the five tools, model selection, credits, and the context cost nobody mentions.

✓ Verified August 2026 OAuth · No API Key · Shared Credits mcp.higgsfield.ai/mcp Setup time: ~5 minutes
Model roster and credit terms change frequently — confirm current details at higgsfield.ai before publishing anything based on this guide.

Part 1 — What It Is

Higgsfield runs a hosted Model Context Protocol server. Connect it to an MCP-capable AI client and that client can generate images, generate video, train reusable characters, and check job status directly in the conversation — without opening the Higgsfield web app.

Authentication is OAuth against your existing Higgsfield account. No API key to manage. No package to install. Your existing plan credits transfer automatically.

The one-sentence version: your AI client's conversation thread becomes your Higgsfield production environment — 30+ models, one connection, one credit pool, no context switching.

Which clients work

Any client that speaks the MCP standard can connect. Confirmed as of August 2026:

✓ Connects via MCP server

  • Claude (web, desktop, mobile)
  • Claude Code
  • Cursor, Windsurf, Cline, OpenCode
  • Cowork, OpenClaw, Hermes Agent, NemoClaw

⚠ Different route

  • ChatGPT — connects via plugin, not MCP server
  • ChatGPT desktop — did not work as of May 2026

All routes use the same Higgsfield account and credit pool.

Part 2 — Setup

Claude — web or desktop

Claude Code and terminal agents

Skip the browser flow. One command adds the connector:

claude mcp add higgsfield <server-url>

The OAuth handshake happens in the browser on first use. Check current syntax in your client documentation — MCP tooling has changed as the standard has matured.

Community vs official: the GitHub repo geopopos/higgsfield_ai_mcp is a community-built local MCP server that requires HIGGSFIELD_API_KEY and HIGGSFIELD_SECRET environment variables and a local Python install. That is a different product from the official hosted OAuth connector documented here.

Part 3 — The Five Tools

The tool surface is small — which is a good thing. Five tools cover the entire platform. Every MCP-capable client calls the same five.

generate_image
Text to image across 16+ models including GPT Image 2, Soul V2, Flux 2, and Nano Banana Pro. Images up to 4K resolution.
generate_video
Images or text prompts into video clips up to 15 seconds. Models include Veo 3.1, Kling 3.0, Sora 2, Seedance 2.0, and MiniMax Hailuo. This is the credit-intensive tool — disable audio passes until the picture is close to final.
create_character
Train a reusable Soul ID from reference images. Keeps a person, model, or product visually consistent across a campaign without re-prompting identity on every generation.
get_generation_status
Poll job progress. Video jobs take 90 seconds or more — this tool keeps the agent in the loop rather than blocking or timing out. In practice your agent kicks off a generation, polls, and returns the finished file in the same conversation.
list_characters
Returns the Soul IDs you've trained. Use this to reference a saved character by name in a generate_video call rather than re-uploading references each session.
Why five tools is the right answer: a larger tool surface means more context tokens loaded upfront and more decision overhead for the agent. Higgsfield's five tools map cleanly to five distinct jobs. Everything else — model selection, aspect ratio, duration — is handled via parameters inside the calls.

Part 4 — The Model Roster

30+ models are available through the connector. The agent selects automatically based on the task, or you can name a model explicitly in your prompt. Naming is worth doing once you know your production preferences — auto-selection is fine for exploration and less predictable for production work.

The seven most commonly used

Character
Soul V2
Character consistency. Lock a face, product, or persona across multiple generations. The foundation of any campaign-length workflow.
Video
Veo 3.1
Google's video model. Strong for photorealistic motion and natural movement. High credit cost per generation.
Video
Kling 3.0
Multi-shot sequences, cinematic camera moves. Good balance of quality and credit cost. See the Kling 3.0 filmmaking guide for prompt workflow.
Image
GPT Image 2
Photographic realism. Best for product shots, portraits, and any image that needs to pass as photographed.
Image
Flux 2
Illustration and stylised output. Strong prompt adherence. Better than photorealistic models when you want a consistent illustrative look.
Image
Nano Banana Pro
Fast iteration. Lower credit cost. Use this for composition exploration and start-frame testing before committing to a high-cost video generation.
Video
Seedance 2.0
Short clips. Efficient for sub-10-second generations. Good for social content that doesn't require cinematic camera language.

Also in the roster

Sora 2, Wan 2.6, MiniMax Hailuo, Soul Cinema, Seedream, Flux Kontext, and 20+ additional image and video models. The full current list is at higgsfield.ai/ai-image and higgsfield.ai/ai-video.

Model naming in prompts: specify the model by the exact name Higgsfield uses in their docs — not colloquial variations. "Use Kling 3.0" works. "Use the Kling model" may resolve to a different version than you expect.

Part 5 — The Context Cost Nobody Mentions

The Higgsfield connector loads 30+ tool definitions into your context window when connected. This is the MCP tax — it's real and it compounds on long conversations.

Context window consumption — before you type anything

Without tool search
~12,000 tokens
With tool search
~few hundred tokens

Source: claudefa.st. Figures are approximate and depend on client implementation.

Practical implications

  • If you run only Higgsfield, this barely matters
  • If you run Higgsfield alongside several other MCP servers, enable tool search or expect a meaningful context tax on every conversation
  • Disconnect connectors you aren't using in a given session
  • On long conversations, the upfront token cost compounds — it's paid out of the context window you'd otherwise use for the work itself
Tool search: when enabled, MCP tool schemas load on demand rather than all at once. This is the correct configuration for anyone running multiple connectors. Check your client's MCP settings for a "tool search" or "lazy loading" option.

Part 6 — Credits and Cost

Free tier (as of May 2026)
~150 credits/month
Authentication
OAuth — no card to start
Credit pool
Shared across all clients
Video cap
15 seconds per generation
Verify before building. Credit allowances, free tier limits, and plan pricing change frequently at AI platforms. The 150 credit figure was reported as of May 2026 and is subject to change. Confirm current allowances at higgsfield.ai before designing a production workflow around a specific credit budget.

Credit efficiency tips

  • Use Nano Banana Pro for start-frame composition testing — lower cost, fast results
  • Generate composition and blocking passes without audio enabled
  • Enable audio only once the picture is close to final
  • Use get_generation_status to poll rather than re-triggering generations
  • Disconnect the connector between sessions if you're running multiple MCP servers

Part 7 — What It Is Good For

Higgsfield MCP wins on aggregation and single authentication. Thirty-plus models, one connector, one credit pool, one polling pattern. The time saved on context switching between platforms is the entire ROI.

✓ Good fits

  • Multi-model workflows in a single conversation
  • Character-consistent campaign content via Soul ID
  • Rapid iteration across image and video without tab switching
  • Agentic workflows where the AI selects and chains tools
  • Teams sharing one credit pool across multiple clients

✗ Not designed for

  • Editing — you get generated assets, not a timeline
  • Long-form video — 15 seconds is the per-generation cap
  • Publishing — distribution is still your job
  • Guaranteed usable output — prompting still matters

Part 8 — Two Things to Get Right Before Publishing

AI disclosure

Platform rules on labeling synthetic media are tightening. EU AI Act Article 50 applies from August 2026. Check the disclosure requirements for where the content will be published — not where it was generated. The requirement follows the destination, not the tool.

Rights and usage

Confirm Higgsfield's current terms on commercial usage and ownership of generated output before building a business on it. Terms vary between platforms and change. Read the current terms of service rather than assuming they match what you've read in a guide — including this one.

Volume without judgment

MCP connectors make generation fast enough that volume stops being the constraint. Judgment becomes the constraint. An agentic workflow that generates 50 clips unattended is only useful if someone reviews all 50 before any go out. Speed of generation does not reduce the need for editorial review — it increases it.

Part 9 — Troubleshooting and Quick Reference

SymptomLikely causeFix
Connected but getting no responseOAuth handshake didn't completeSettings → Connectors → disconnect Higgsfield → reconnect → complete the browser auth window fully
Models listed but generations failingInsufficient creditsCheck balance at higgsfield.ai dashboard. Top up or wait for monthly reset
Context window filling fast30+ tool definitions loaded without tool searchEnable tool search in client MCP settings. Disconnect unused connectors between sessions
Agent keeps re-triggering generationsNot using get_generation_status to pollExplicitly ask the agent to poll job status rather than re-generate. Video jobs take 90+ seconds
URL not connectingServer URL has changedRe-check higgsfield.ai/mcp for the current URL. Remove the connector and add fresh with the new URL
ChatGPT not finding the MCP serverChatGPT uses the plugin, not MCP serverAdd Higgsfield via ChatGPT's plugin directory, not via MCP connector settings
Wrong model being selectedAuto-selection chose a different model than expectedName the model explicitly in your prompt: "Use Kling 3.0 for this generation"

Quick reference

MCP server URL
mcp.higgsfield.ai/mcp
Auth method
OAuth (no API key)
Tools
5 (generate_image, generate_video, create_character, get_generation_status, list_characters)
Max video length
15 seconds per generation
Image resolution
Up to 4K
Verify connection
List my available Higgsfield models and show my current credit balance

Sources

  • higgsfield.ai/mcp — official MCP documentation
  • Higgsfield blog — "How To Generate AI Videos Straight From Claude with Higgsfield's MCP"
  • Higgsfield blog — "MCP for Marketers"
  • claudefa.st — "Higgsfield MCP: Sora, Veo, Kling from Claude Code"
  • TECHSY — "Higgsfield MCP for Claude Code: 60-Second Setup"
  • Jonathans Blog — "How To Use Higgsfield With Claude (via MCP)"
  • higgsfieldmcp.com — independent resource, not affiliated with Higgsfield
  • GitHub: geopopos/higgsfield_ai_mcp — community-built local server (different product)

Related Guides

Kling 3.0 AI Filmmaking GuideMulti-shot storyboarding, Elements character locking, native audio — available through Higgsfield MCP
AI Video Prompt GuideThe shot brief framework — works across Kling, Veo, Sora, and every model in the Higgsfield roster
AI Video From PhotoMotion prompt templates for image-to-video — directly applicable to Higgsfield's generate_video tool
Revid AI ReviewScript-to-video workflow — an alternative approach to Higgsfield for faceless content and social video

Frequently Asked Questions

What is the Higgsfield MCP server URL?
The current server URL is https://mcp.higgsfield.ai/mcp. Some sources list it without the /mcp path. Confirm the current URL at higgsfield.ai/mcp before connecting — MCP infrastructure URLs change as platforms update their implementations.
Do you need an API key for Higgsfield MCP?
No. The official hosted connector uses OAuth against your existing Higgsfield account — no API key to generate or manage. The community-built GitHub server (geopopos/higgsfield_ai_mcp) does require API keys but that is a separate, unofficial product.
Which AI clients work with Higgsfield MCP?
Any MCP-capable client. Confirmed: Claude (web, desktop, mobile, Claude Code), Cursor, Windsurf, Cline, OpenCode, Cowork, OpenClaw, Hermes Agent, NemoClaw. ChatGPT connects via a plugin rather than the MCP server. All share the same Higgsfield account and credit pool.
How many credits does Higgsfield MCP use?
Credit costs vary by model and generation type. The free tier provided approximately 150 credits per month as of May 2026 — verify current allowances at higgsfield.ai before building a production workflow around a specific budget.
What is the context cost of the Higgsfield MCP connector?
Without tool search enabled, 30+ tool definitions load upfront and can consume up to 12,000 tokens before you type anything. With tool search enabled, schemas load on demand and the cost drops to a few hundred tokens. If you run multiple MCP connectors, enable tool search.