Skip to content
All releases
Added

SubQ Savant 1.4

Subquadratic's 4M-token reasoning model joins the catalog with platform and BYOK routes.

SubQ Savant 1.4 is Subquadratic's long-context reasoning model. It reads up to 4M tokens in one request, reasons before it answers, and accepts text, images and files on OpenAI-compatible Chat Completions and Responses endpoints. It costs $0.10 per 1M input tokens, $0.01 per 1M cached input tokens and $0.50 per 1M output tokens, with reasoning billed as output.

curl
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "subquadratic/subq-savant-1.4",
    "messages": [{"role": "user", "content": "Summarize the risk clauses in this contract set."}]
  }'
AI SDK
import { createAnyRouter } from "@anyr/ai-sdk-provider"
import { generateText } from "ai"
const anyrouter = createAnyRouter()
const { text } = await generateText({ model: anyrouter("subquadratic/subq-savant-1.4"), prompt: "Hello" })
Coding agent
curl -fsSL https://anyrouter.dev/cli.sh | bash
anyr login
anyr claude --model "subquadratic/subq-savant-1.4"

Reasoning runs at high effort by default; send low or max to change it, and omit temperature because it is fixed at 1. Savant is in early access, so your own SubQ key must be enabled for the model. Read the SubQ model docs and release notes for the full claims.

Playbooks: AI SDK, coding agents, OpenAI SDK.

Looking for the full catalog? Browse all models.