Ultra-budget open-model inference
Need the cheapest OpenAI-compatible host for high-volume open weights without caring about brand prestige.
At a glance — Oct 2026
Updated 2026-10-27 · forward this page: /jtbds/budget-open-inference#briefScores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.
- Depot Pick
- DeepInfra · 7.95/10 for this job (curated estimate, high confidence)
Overweight pricePerf. DeepInfra wins catalog+price; Novita/SiliconFlow/Hyperbolic as pure price peers.
- Runner-up
- Novita AI · 7.42/10
Scores higher than the pick on: Privacy / data retention.
- When not to use the pick
- Need Fireworks-level production features/tool calling reliability, or compliance blocks smaller hosts.
- Lock-in / migration
- The pick has lower stored lock-in risk than runner-up Novita AI (3 vs 3.5). Pick exit signals: serves open-weight models (provider tag `open-models`); OpenAI-compatible API (provider tag `openai-compat`). Self-host fallback listed for this job: Qwen3-0.6B-Base, DeepSeek-V3.2 weights, Ollama.
- Sourced price (usd_per_million_input_tokens)
Pick's lowest sourced SKU: Llama-3.3-70B-Instruct-Turbo $0.1 (usd_per_million_input_tokens) — source, as of 2026-10-03, re-checked 2026-10-27
Cheapest matching SKU: Chutes Qwen3.8-27B (Qwen/Qwen3.8-27B-TEE) $0.039 — source, as of 2026-10-27, re-checked 2026-10-27. Not the pick; not quality-adjusted.
Price history: first seen 2026-10-03 · last changed 2026-10-19 · history JSON
- Free option
- Failure modes to watch
- Runaway spend: Set a budget and a tool-call cap per task. Alert when spend jumps. Stop the run at the cap instead of summarizing a partial trace as success.
Same-job, same-unit comparisons: DeepInfra vs Chutes
Cite: Agent Provider Depot, Depot Pick for budget-open-inference: deepinfra, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=budget-open-inference
Free Ultra-budget open-model inference options · All sourced prices, cheapest first
DeepInfra
Overweight pricePerf. DeepInfra wins catalog+price; Novita/SiliconFlow/Hyperbolic as pure price peers.
Pricing (directional): As of 2026-09-14: Broad OpenAI-compatible catalog with model-specific token rates (e.g. Llama-3.3-70B Turbo $0.10/$0.32 per 1M in/out; DeepSeek-V4-Flash input $0.09, cached input $0.018 and output $0.18 per 1M (deepinfra.com/pricing as of 2…
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Budget open-model workhorse and one of the cheapest large-catalog hosts. Great for evals and high-volume non-critical traffic; dual-run critical paths on Fireworks/Together.
Quote for Ultra-budget open-model inference
Depot Pick stays DeepInfra. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=budget-open-inference
Pick is not a price winner. deepinfra sourced input $/1M is 0.1; cheapest among candidates is chutes at 0.039.
- bge-base-en-v1.5 usd_per_million_input_tokens $0.005 source as of 2026-10-03
- Llama-3.3-70B-Instruct-Turbo usd_per_million_input_tokens $0.1 source as of 2026-10-03
- Llama-3.3-70B-Instruct-Turbo usd_per_million_output_tokens $0.32 source as of 2026-10-03
- A100 usd_per_gpu_hour $1.49 source as of 2026-10-20
- B200 usd_per_gpu_hour $4.89 source as of 2026-10-20
- B300 usd_per_gpu_hour $5.89 source as of 2026-10-20
- H100 usd_per_gpu_hour $2.9 source as of 2026-10-20
- H200 usd_per_gpu_hour $3.49 source as of 2026-10-20
4 of 8 candidates have no sourced price row. 4 are excluded from the Input $/1M comparison (chat_completion_input). min sourced input SKU among candidates; not quality-adjusted
Sourced Input $/1M for Ultra-budget open-model inference
Candidates for this job that already have a matching sourced price (chat_completion_input), cheapest first. As of 2026-10-27. Each figure is one SKU in that unit, not a blend of input and output, not a cross-modal rate, and not a Depot score. min sourced input SKU among candidates; not quality-adjusted The Depot Pick is not a price winner.
| Provider | Input $/1M | SKU | Source | As of |
|---|---|---|---|---|
| Chutes | $0.039 | Qwen3.8-27B (Qwen/Qwen3.8-27B-TEE) | pricing page | 2026-10-27 |
| Z.AI (Zhipu GLM) | $0.04 | GLM-4.6V-FlashX | pricing page | 2026-10-03 |
| DeepInfraDepot Pick | $0.1 | Llama-3.3-70B-Instruct-Turbo | pricing page | 2026-10-03 |
| MiniMax | $0.15 | MiniMax-M3.1-Flash | pricing page | 2026-10-19 |
Runners-up
Novita AI
Budget open-model host with a pinned official matrix — strong DeepInfra/SiliconFlow peer. Prefer…
SiliconFlow
Price/catalog monster for open models. Confirm data-residency/compliance before production.
Hyperbolic
Cheap open-model inference peer to Novita/DeepInfra. Good for evals and non-critical volume.
Z.AI (Zhipu GLM)
International GLM via z.ai — catalog zai, not a second Zhipu mainland-only SKU. Strong cheap/coding runner…
Chutes
Budget open-model overflow with a now-pinned public TEE matrix. Depot open-model production pick stays…
MiniMax
Primary indie-usable Chinese multimodal API alongside Moonshot. Prefer Anthropic/OpenAI for coding agents…
When NOT to use the top pick
Need Fireworks-level production features/tool calling reliability, or compliance blocks smaller hosts.
Full ranking (score table)
Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.
| # | Provider | Overall | Fit | $/perf | Latency | DX | As of |
|---|---|---|---|---|---|---|---|
| 1 | IONOS AI Model Hub | 8.0 | 7.8 | 8.4 | 8.0 | 7.8 | 2026-09-14 |
| 2 | DeepInfraDepot Pick | 8.0 | 8.0 | 9.4 | 7.5 | 8.5 | 2026-09-14 |
| 3 | MiniMax | 7.8 | 8.4 | 8.8 | 7.8 | 8.2 | 2026-09-14 |
| 4 | Z.AI (Zhipu GLM) | 7.8 | 8.2 | 9.0 | 7.6 | 8.0 | 2026-09-14 |
| 5 | Novita AI | 7.4 | 7.2 | 9.0 | 7.0 | 7.8 | 2026-09-14 |
| 6 | SiliconFlow | 7.3 | 7.4 | 9.2 | 6.8 | 7.6 | 2026-09-14 |
| 7 | Hyperbolic | 7.3 | 7.0 | 8.8 | 6.8 | 8.0 | 2026-09-14 |
| 8 | Chutes | 7.3 | 7.4 | 9.2 | 6.8 | 7.6 | 2026-09-14 |
Score methodology for this job
Overweight pricePerf. DeepInfra wins catalog+price; Novita/SiliconFlow/Hyperbolic as pure price peers.
capabilityFit: 25%pricePerformance: 18%latencyReliability: 15%developerDx: 15%privacyPosture: 10%lockInRisk: 7%maturity: 10%