Browserbase — Free 3 concurrent / 1 browser hour / 1k Search+Fetch public / non-affiliate
Sign up on Browserbase Free plan. Stay within concurrent/hour/Search+Fetch caps.
Browserless — Free 1k units/mo / 2 concurrent public / non-affiliate
Sign up for Browserless Free plan. Stay within 1k units/mo and 2 concurrent browsers.
GitHub Codespaces — Free 120 core-hrs + 15 GB-mo (personal) public / non-affiliate
Use Codespaces on a personal GitHub Free account within included monthly compute/storage. Org/enterprise plans do not include a free Codespaces quota per docs.
Groq — Free plan LLM inference (rate-limited) public / non-affiliate
Create a Groq account and use Free plan rate limits for agent LLM calls. Check current Free Plan Limits table per model; upgrade if you hit 429s.
Hyperbrowser — Free 5,000 credits public / non-affiliate
Sign up at Hyperbrowser. Use free 5,000 credits; confirm current free allowance on pricing.
Kernel — Free $5 usage/mo (refreshed) public / non-affiliate
Sign up at Kernel pricing page. Use monthly free $5 allotment; confirm stealth/free limits on current pricing.
Modal — $30/month free compute credit (Starter) public / non-affiliate
Sign up for Modal Starter. Use the $30/month free compute credit for GPU/CPU jobs. Confirm current allotment on pricing before scaling.
Steel — $30 one-time free browser usage public / non-affiliate
Sign up at Steel. Use free $30 usage within 90 days. Captcha/proxy features may require a $10 deposit — confirm on docs before enabling.
99Stack Cloud — hourly-billed cloud VPS and GPU instances (H100, L40S, A40, etc.) (referral credits) New affiliate
Updated Oct 25, 2026
Deploy hourly-billed cloud VPS or GPU instances via 99Stack Cloud's redeem_url (pre-paid balance). Delete instances when idle to avoid charges; confirm current rates at launch.
GigaGPU — dedicated bare-metal GPU servers, UK datacenter (affiliate) New affiliate
Updated Oct 24, 2026
Order a dedicated GPU server (self-hosted LLMs, inference, fine-tuning, rendering) via GigaGPU's redeem_url; open it in a browser if a scripted request hits Cloudflare. Confirm current monthly price and specs at checkout before ordering.
Novita AI — GPU inference API (affiliate) affiliate
Sign up via the redeem_url for Novita AI.
RunPod — GPU compute credit referral affiliate
Sign up via the redeem_url for RunPod GPU/Serverless compute. Suited to agent workloads needing GPUs.
SiliconFlow China (siliconflow.cn, Chinese-language site) — LLM inference API (affiliate) New affiliate
Updated Oct 24, 2026
Use only if the operator can use a Chinese-language site and likely a mainland China account and payment method: pay-as-you-go LLM/multimodal inference API (open models like DeepSeek, Qwen, GLM) priced in CNY via SiliconFlow China's redeem_url (siliconflow.cn, not siliconflow.com).
Thunder Compute — low-cost on-demand GPU cloud (A6000, L40, A100, H100) (referral credits) New affiliate
Updated Oct 25, 2026
Launch on-demand GPU instances (A6000, L40, A100, H100) for training or inference via Thunder Compute's redeem_url. Stop or delete instances when idle to avoid charges; confirm current rates at launch.
Vast.ai — GPU cloud (affiliate) affiliate
Sign up via the redeem_url for Vast.ai GPU cloud.
WaveSpeedAI — pay-as-you-go inference API for image, video and audio generative models (affiliate) New affiliate
Updated Oct 24, 2026
Generate images, video and audio via WaveSpeedAI's API, CLI or MCP through the redeem_url. Check per-model pricing before batch jobs; follow model licenses and content policies (no deceptive deepfakes or non-consensual likenesses).