Nova 2 Lite is a fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text.
Nova 2 Lite demonstrates standout capabilities in processing documents, extracting information from videos, generating code, providing accurate grounded answers, and automating multi-step agentic workflows.
Modalities
In / Out Price
$0.30 / $2.50per 1M
Context
1.0M
Released
Dec 2, 2025
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), or Exacto (highest tool-calling accuracy).
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Scores on standardized evaluations. Higher percentages are better — and rank percentile shows where this model lands among all models on OpenRouter.
| Source | Benchmark | Score |
|---|---|---|
| Artificial Analysis | Nova 2.0 Lite (high) Coding Index | 23.0 |
| Artificial Analysis | Nova 2.0 Lite (high) GPQA Diamond | 81.1% |
| Artificial Analysis | Nova 2.0 Lite (high) HLE | 11.6% |
| Artificial Analysis | Nova 2.0 Lite (high) IFBench | 70.7% |
| Artificial Analysis | Nova 2.0 Lite (high) τ²-Bench Telecom | 72.8% |
| Artificial Analysis | Nova 2.0 Lite (high) AA-LCR | 60.3% |
| Artificial Analysis | Nova 2.0 Lite (high) GDPval-AA | 1.8% |
| Artificial Analysis | Nova 2.0 Lite (high) CritPt | 0.3% |
| Artificial Analysis | Nova 2.0 Lite (high) Terminal-Bench Hard | 16.7% |
| Artificial Analysis | Nova 2.0 Lite (high) AA-Omniscience Accuracy | 19.0% |
| Artificial Analysis | Nova 2.0 Lite (high) AA-Omniscience Non-Hallucination Rate | 8.6% |
| Artificial Analysis | Nova 2.0 Lite (medium) GPQA Diamond | 76.8% |
| Artificial Analysis | Nova 2.0 Lite (medium) HLE | 9.0% |
| Artificial Analysis | Nova 2.0 Lite (medium) IFBench | 68.5% |
| Artificial Analysis | Nova 2.0 Lite (medium) τ²-Bench Telecom | 75.7% |
| Artificial Analysis | Nova 2.0 Lite (medium) AA-LCR | 60.0% |
| Artificial Analysis | Nova 2.0 Lite (medium) CritPt | 0.0% |
| Artificial Analysis | Nova 2.0 Lite (medium) Terminal-Bench Hard | 17.4% |
| Artificial Analysis | Nova 2.0 Lite (medium) AA-Omniscience Accuracy | 18.2% |
| Artificial Analysis | Nova 2.0 Lite (medium) AA-Omniscience Non-Hallucination Rate | 8.2% |
| Artificial Analysis | Nova 2.0 Lite (low) GPQA Diamond | 69.8% |
| Artificial Analysis | Nova 2.0 Lite (low) HLE | 4.0% |
| Artificial Analysis | Nova 2.0 Lite (low) IFBench | 61.2% |
| Artificial Analysis | Nova 2.0 Lite (low) τ²-Bench Telecom | 71.9% |
| Artificial Analysis | Nova 2.0 Lite (low) AA-LCR | 54.3% |
| Artificial Analysis | Nova 2.0 Lite (low) CritPt | 0.0% |
| Artificial Analysis | Nova 2.0 Lite (low) Terminal-Bench Hard | 3.8% |
| Artificial Analysis | Nova 2.0 Lite (low) AA-Omniscience Accuracy | 17.6% |
| Artificial Analysis | Nova 2.0 Lite (low) AA-Omniscience Non-Hallucination Rate | 12.9% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) GPQA Diamond | 60.3% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) HLE | 2.9% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) IFBench | 40.5% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) τ²-Bench Telecom | 62.0% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) AA-LCR | 18.7% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) CritPt | 0.0% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) Terminal-Bench Hard | 6.8% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) AA-Omniscience Accuracy | 14.0% |
| Artificial Analysis | Nova 2.0 Lite (Non-reasoning) AA-Omniscience Non-Hallucination Rate | 14.1% |
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Token volume and request traffic to this model over time.
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
| $0.30 | $2.50 | 0.81s | 15 tps | ||
| $0.30 | $2.50 | 0.71s | 80 tps |
Throughput
80tok/s
P50, best across providers
Latency
0.71s
P50, best provider
100.00%
99.99%
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.
Nova 2 Lite is a fast, cost-effective reasoning model for everyday workloads that can process text, images, and videos to generate text. Nova 2 Lite demonstrates standout capabilities in processing documents, extracting information from videos, generating code, providing accurate grounded answers, and automating multi-step agentic workflows.
Nova 2 Lite costs $0.30/M input tokens and $2.50/M output tokens.
Nova 2 Lite has a 1,000,000 token context window. It supports up to 65,535 completion tokens.
Yes. Nova 2 Lite accepts tools and tool_choice for function calling. It does not support response_format, so JSON output is not enforced.
Nova 2 Lite accepts text, images, video and files such as PDFs as input and returns text.
Nova Premier 1.0, Nova Lite 1.0, Nova Micro 1.0 and 1 more are other text models from Amazon.
Nova 2 Lite was released on December 2, 2025.
