All Models - Corvic Platform

Corvic offers 45 models from eight providers. Use them in chats, agents, and pipelines, or call them from your own code through Corvic’s OpenAI-compatible gateway with one API key and one bill. The same list, with the same details, is under Settings → Models in the app. Text 33 · Classification 1 · Voice 2 · Embedding 7 · Web search 2 Ready to call one? Jump to Start using a model.

Text

Chat, agents, and LLM augment.

Corvic4 models

Corvic’s own tiers, tuned for data analysis; see Corvic models. Auto picks a tier per message, so it has no API ID.

Model	API model ID	Context	Modalities	Price	Trains on data
Auto	—	1M / 128K	Text, Image → Text	$$	No
Corvic Jackdaw	`corvic/jackdaw`	1M / 128K	Text, Image → Text	$	No
Corvic Rook	`corvic/rook`	1M / 128K	Text, Image → Text	$	No
Corvic Raven	`corvic/raven`	1M / 128K	Text, Image → Text	$	No

Anthropic8 models

Model	API model ID	Context	Modalities	Price	Trains on data
Claude Opus 5.5	`claude-opus-5-5`	1M / 128K	Text, Image → Text	$$	No
Claude Sonnet 5.5	`claude-sonnet-5-5`	1M / 128K	Text, Image → Text	$	No
Claude Haiku 5.5	`claude-haiku-5-5`	1M / 128K	Text, Image → Text	$	No
Claude Fable 5	`claude-fable-5`	1M / 128K	Text, Image → Text	$$$	No
Claude Opus 5	`claude-opus-5`	1M / 128K	Text, Image → Text	$$	No
Claude Sonnet 5	`claude-sonnet-5`	1M / 128K	Text, Image → Text	$$	No
Claude Opus 4.8	`claude-opus-4-8`	1M / 128K	Text, Image → Text	$$	No
Claude Sonnet 4.6	`claude-sonnet-4-6`	1M / 128K	Text, Image → Text	$$	No

OpenAI9 models

Model	API model ID	Context	Modalities	Price	Trains on data
GPT 6.1 Sol	`gpt-6.1-sol`	1.1M / 128K	Text, Image → Text	$$	No
GPT 6.1 Terra	`gpt-6.1-terra`	1.1M / 128K	Text, Image → Text	$	No
GPT 6 Astra	`gpt-6-astra`	1.1M / 128K	Text, Image → Text	$$$	No
GPT 6 Sol	`gpt-6-sol`	1.1M / 128K	Text, Image → Text	$$	No
GPT 6 Luna	`gpt-6-luna`	1.1M / 128K	Text, Image → Text	$	No
GPT 5.6 Sol	`gpt-5.6-sol`	1.1M / 128K	Text, Image → Text	$$$	No
GPT 5.6 Terra	`gpt-5.6-terra`	1.1M / 128K	Text, Image → Text	$$	No
GPT 5.6 Luna	`gpt-5.6-luna`	1.1M / 128K	Text, Image → Text	$	No
GPT 5.4	`gpt-5.4-2026-03-05`	1.1M / 128K	Text, Image → Text	$$	No

xAI3 models

Model	API model ID	Context	Modalities	Price	Trains on data
Grok 4.7	`grok-4.7`	500K / 500K	Text, Image → Text	$	No
Grok 4.6	`grok-4.6`	500K / 500K	Text, Image → Text	$	No
Grok 4.5	`grok-4.5`	500K / 500K	Text, Image → Text	$	No

Google3 models

Model	API model ID	Context	Modalities	Price	Trains on data
Gemini 4 Argon	`gemini-4-argon`	1.0M / 66K	Text, Image → Text	$$	No
Gemini 3.8 Flash	`gemini-3.8-flash`	1.0M / 66K	Text, Image → Text	$	No
Gemini 3.7 Flash	`gemini-3.7-flash`	1.0M / 66K	Text, Image → Text	$	No

Meta2 models

Model	API model ID	Context	Modalities	Price	Trains on data
Muse Wonder 1.0	`muse-wonder-1.0`	1.0M / 1.0M	Text, Image → Text	$$	No
Muse Spark 1.3	`muse-spark-1.3`	1.0M / 1.0M	Text, Image → Text	$	No

OpenRouter4 models

Open-weight models (Moonshot Kimi and Z.ai GLM) served through OpenRouter.

Model	API model ID	Context	Modalities	Price	Trains on data
Kimi K3.1 Pro	`moonshotai/kimi-k3.1-pro`	1.0M / 944K	Text, Image → Text	$	No
Kimi K3	`moonshotai/kimi-k3`	1.0M / 1.0M	Text, Image → Text	$$	No
GLM 5.3 Flash	`z-ai/glm-5.3-flash`	1.0M / 131K	Text, Image → Text	$	No
GLM 5.2	`z-ai/glm-5.2`	1.0M / 131K	Text → Text	$	No

Classification

Decisions, scores, and labels.

TypeSafe AI1 model

A decision model: text in, a choice, score, or yes/no out. Call it from pipelines or code at the router’s /typesafe/v1/systemone endpoint, not chat completions.

Model	API model ID	Context	Modalities	Price	Trains on data
Jev	`jev-1.13.0`	32K	Text → Decision	—	No

Voice

Speech to text.

OpenAI2 models

Speech-to-text behind voice dictation, billed per minute of audio.

Model	API model ID	Context	Modalities	Price	Trains on data
GPT Transcribe	—	—	Audio → Text	$/min	No
GPT Live Transcribe	—	—	Audio → Text	$$/min	No

Embedding

Text and image embeddings for spaces and search.

OpenAI3 models

Text embedders are used by embedding actions in pipelines; CLIP embeds images.

Model	API model ID	Context	Modalities	Price	Trains on data
Text Embedding 3 Small	—	8K	Text → Embeddings	—	No
Text Embedding 3 Large	—	8K	Text → Embeddings	—	No
CLIP	—	—	Image → Embeddings	—	No

Google4 models

SigLIP 2 embeds images; the others embed text.

Model	API model ID	Context	Modalities	Price	Trains on data
Text Embedding 004	—	2K	Text → Embeddings	—	No
Gemini Embedding	—	2K	Text → Embeddings	—	No
Gemini Embedding 2	—	8K	Text → Embeddings	—	No
SigLIP 2	—	—	Image → Embeddings	—	No

Web search

Answers and pages from the live web.

Corvic2 models

Billed per request. Web Augment answers each row from the live web (call it at /responses; reasoning.effort sets how deep it looks). Web Search finds pages for agents. The search vendor publishes no retention terms, so these are marked as possibly training on data.

Model	API model ID	Context	Modalities	Price	Trains on data
Corvic Web Augment	`corvic/web-augment`	32K / 4K	Text → Text	$/req	Yes
Corvic Web Search	—	—	Text → Text	$/req	Yes

Start using a model

Create an API key

In Corvic, open Settings → Models, select a model, and choose Create API key under Use via API. Copy the key; it’s shown once and lasts 90 days.

Set it in your environment

export CORVIC_API_KEY="paste-your-key"

Call any model by its API model ID

Point any OpenAI SDK at https://api.corvic.ai/openai/v1 and pass an ID from the tables below.

curl https://api.corvic.ai/openai/v1/chat/completions \
  -H "Authorization: Bearer $CORVIC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "corvic/jackdaw",
    "messages": [{"role": "user", "content": "Hello, Corvic"}]
  }'

import os
from openai import OpenAI

client = OpenAI(base_url="https://api.corvic.ai/openai/v1", api_key=os.environ["CORVIC_API_KEY"])
reply = client.chat.completions.create(
    model="claude-sonnet-5-5",
    messages=[{"role": "user", "content": "Hello, Corvic"}],
)
print(reply.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.corvic.ai/openai/v1", apiKey: process.env.CORVIC_API_KEY });
const reply = await client.chat.completions.create({
  model: "gpt-6.1-terra",
  messages: [{ role: "user", content: "Hello, Corvic" }],
});
console.log(reply.choices[0].message.content);

GET /models lists the IDs your organization can call. See the API Playground for every endpoint, streaming, and reasoning effort.

Reading the tables

Column	Meaning
API model ID	The `model` value for the gateway. A dash means the model runs in Corvic’s chats, agents, and pipelines but isn’t served over the API.
Context	Input / output: the most tokens a request can carry in, and the most a reply can return.
Modalities	What the model takes in → what it gives back.
Price	Relative cost of output: $ low, $$ medium, $$$ high. Per request (/req) or per minute of audio (/min) where the model is billed that way. Usage is billed in AI tokens; see Settings → Usage.
Trains on data	Whether the provider may train on your prompts and completions. No means the provider does not.

Bring your own model

Admins can add their own endpoints under Settings → Models → Bring your own model. Models you add appear in the model picker and this list, and you call them over the API by the ID Corvic gives them.

Choose the API type

OpenAI, OpenAI Completions, OpenAI Responses, Anthropic, Anthropic Messages, Azure OpenAI, Gemini, Groq, Hugging Face, Databricks, OpenRouter, Together AI, Ollama, LM Studio, Keyword AI, or Unify AI.

Enter the endpoint

Fill in Model, Base URL, and API Key (for Azure OpenAI, also API Version and Deployment ID). Endpoint names must be unique.

Validate and create

Choose Validate to check the endpoint, then Create. An endpoint can be saved without validating; it then shows Saved, but not validated.

Admins can limit which models rooms may use with a policy.

​Text

​Classification

​Voice

​Embedding

​Web search

​Start using a model

​Reading the tables

​Bring your own model