Voxtral Mini 3B (Audio)
Voxtral Mini 3B (Audio) is a chat model made by Mistral AI, served on nRouter.
Mistral AI
AWS Bedrock
Start building with voxtral-mini-3b-2507
Get a key
Set your API key
export NROUTER_API_KEY="<YOUR_API_KEY>"Configure Claude Code
# Point Claude Code to nRouter gateway
export ANTHROPIC_BASE_URL="https://api.nrouter.ai"
export ANTHROPIC_API_KEY="$NROUTER_API_KEY"
# Launch with bedrock/mistral.voxtral-mini-3b-2507
claude --model bedrock/mistral.voxtral-mini-3b-2507Pricing
Rates are read live
Pricing for voxtral-mini-3b-2507 is fetched from the live catalogue on load, so it is never served from a cache that could outlive a repricing.
Pricing Transparency & Zero-Markup Guarantee
nRouter strictly operates with a Zero per-token markup guarantee. All inference requests for Voxtral Mini 3B (Audio) are billed at exact upstream provider list prices with zero per-token margin, zero routing fees, and zero hidden platform overhead.
Token consumption is settled at the published upstream rate of the exact served model and provider tier that executed your call.
Upstream prompt caching discounts (up to 50%–90% reduction on cached prompt tokens) pass through directly to your balance with zero added fee.
Every response includes the authoritative x-nr-request-cost header for instant accounting and FinOps reconciliation.
Specifications
—
—
chat
Context window, compared
Bars are scaled to 131.1K tokens. A model with no published context window reads Unknown, never zero.
Context Limits & Caching Economics
With an effective context ceiling of — tokens and a maximum output generation ceiling of — tokens, voxtral-mini-3b-2507 is engineered to support long-document analysis, repository code exploration, and complex multi-turn dialog without token truncation.
Upstream prefix caching stores static context (system prompts, tool definitions, documentation corpora) in memory. Cached token reads receive up to 50%–90% cost discounts from upstream providers, passed through directly to your nRouter balance with zero added fee.
- ✦Autonomous Tool Use: Structured JSON schema validation and multi-step function calling loops.
- ✦Code Intelligence: Multi-file codebase refactoring, unit test generation, and complex algorithmic reasoning.
- ✦High-Density RAG: In-context information extraction across large token windows without needle loss.
- ✦Multimodal Ingestion: Visual documentation analysis, image parsing, and structured asset extraction.
Provider-Reported Capabilities
What the hosting provider reports for this model, shown as it was returned. nRouter adds nothing to this report and infers nothing that is missing from it.
mistral.voxtral-mini-3b-2507
null
32768
null
Voxtral Mini 3B 2507
null
Voxtral
Mistral AI
ACTIVE
null
null
null
null
Multi-Provider Redundancy
nRouter routes traffic for Voxtral Mini 3B (Audio) across multiple underlying cloud hyperscalers to guarantee continuous zero-downtime availability and eliminate vendor lock-in:
Traffic for bedrock/mistral.voxtral-mini-3b-2507 routes dynamically across Microsoft Azure, AWS Bedrock, Google Cloud Vertex AI, and direct provider endpoints without code changes.
If an upstream cloud provider reports rate limits (HTTP 429), regional degradation, or an outage, nRouter automatically retries on an alternate cloud backend in <1ms.
A single nRouter virtual key provides unified access with enforced spending budgets, model allowlists, and end-to-end SOC 2 compliant audit logging.
Availability
Not enough data yet
We have not collected enough health probes for this model to publish an uptime or latency figure. A percentage from a handful of samples is fabricated precision, so none is shown until the sample count supports one.
Interactive Playground
Send a real test request to voxtral-mini-3b-2507 using your virtual key.
Test it
curl https://api.nrouter.ai/v1/chat/completions \
-H "Authorization: Bearer $NROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"bedrock/mistral.voxtral-mini-3b-2507","messages":[{"role":"user","content":"Hello!"}]}'