One key.
Every model.
Zero rate ceilings.
The AI gateway built for autonomous agents.
One OpenAI-compatible API for leading models, intelligent routing, provider failover, and infrastructure designed to keep agents running.
Already using OpenAI?
Change your base URL and API key.
Keep your streaming, tool-calls, and JSON mode. Migration takes minutes, not days. Every request is observable in the Forge dashboard and billed against a single unified balance.
- base_url
- https://forge-main.pages.dev/v1
- provider proxy
- TopTools (top-tools-ai.com/api/v1)
- auth
- Bearer fg-*
- endpoints
- /chat/completions · /embeddings · /models
- region
- iad1 · sfo1 · fra1 · hkg1
from openai import OpenAI
client = OpenAI(
api_key="fg-demotoken1234567890",
base_url="https://forge-main.pages.dev/v1",
)
response = client.chat.completions.create(
model="deepseek-r1",
messages=[{"role": "user", "content": "Explain load balancing."}],
stream=True,
)
for chunk in response:
print(chunk.choices[0].delta.content, end="")Agents don't make one request.
Autonomous agents create deep, repeated chains of model calls. When an agent loops 10 times to verify a file, a single provider failure or 429 rate limit breaks the entire workflow.
Forge isn't built for simple chatbots. It is designed to keep agentic loops running smoothly by handling failures at the network edge.
Frontier models, one endpoint.
Pay for tokens, not seats.
No forced monthly seat minimums. Transparent usage billing with zero markup on upstream provider tokens.
Developer Pay-As-You-Go
POPULARIdeal for individual developers building agents, scripts, and production apps with transparent token billing.
- Access to all 28+ frontier models (Claude 3.7, GPT-4o, DeepSeek R1)
- Zero monthly seat fees — pay only for tokens you consume
- OpenAI SDK drop-in replacement (base_url: https://forge-main.pages.dev/v1)
- Automatic key rotation & sub-500ms edge failover
- Real-time telemetry & latency metrics dashboard
Pro Scale
RECOMMENDEDBuilt for teams and heavy agent workloads requiring guaranteed high rate limits and dedicated throughput.
- $8.00 included monthly usage credit
- Higher rate limit ceilings across all providers (100k+ TPM)
- Bring-Your-Own-Key (BYOK) fallback integration
- Team workspace & shared API key management
- 99.99% Routing Uptime SLA guarantee
- Priority Discord & Email developer support
Enterprise Custom
ENTERPRISEDedicated routing clusters, zero data retention, and custom SLAs for enterprise AI applications.
- Dedicated isolated edge proxy clusters
- Zero Data Retention (ZDR) & SOC-2 compliance
- Custom SLA guarantees & custom rate limits
- Custom model fine-tuning & private endpoint proxying
- 24/7 Phone & Slack channel support
- Custom invoice & purchase order billing