Inception - 100M Free Tokens for Mercury 2

Overview

Inception provides developer access to Mercury 2 with a substantial free tier designed to let teams benchmark production-scale workloads without immediate infrastructure commitments. Every newly generated Inception API key automatically includes 100 million free tokens (expanded tenfold from the initial 10 million allocation).

Along with the increased token allowance, Inception expanded free-tier throughput by introducing 10x higher rate limits, enabling developers to simulate realistic traffic and test high-volume pipelines without hitting artificial request ceilings.

---

Key Features & Improvements

The free allocation applies to the updated Mercury 2 model, which includes several architectural and inference optimizations:

• Lower Time-To-First-Token (TTFT): Dramatically reduces initial response time, crucial for streaming interactions.
• Reduced End-to-End Latency: Faster complete response generation across long-context queries and complex agent tasks.
• Enhanced Tool Calling: Significantly more reliable function execution and structured output generation.

---

Target Use Cases

Mercury 2's low-latency performance profile is tailored for latency-sensitive applications where overhead compounds across multiple calls:

• Real-time Voice Agents: Interactive spoken conversational systems requiring sub-second turnarounds.
• Search & Retrieval (RAG): High-throughput reranking and query expansion capable of firing dozens of times per search request.
• Coding Agents & Multi-Agent Chains: Multi-step programmatic reasoning and automated subagent workflows.

---

How to Access

1. Visit Inception and sign up for an API account.
2. Generate a new API key to automatically receive 100 million free tokens with enhanced rate limits.
3. Mercury 2 can also be integrated directly through Baseten for teams running workloads on that platform.