Protoface opens its API to 40+ models from $0.003 a second

Protoface said on October 8, 2026 that its API now serves more than 40 generative media models from over 20 providers under one key, priced from $0.003 a second.

The company announced the expansion in a press release distributed by EIN Presswire. Protoface describes itself as a generative media inference platform: one account and one API key for video, image, audio and 3D models, with inference it tunes in-house to cut the cost of running open-weight models.

By Protoface's own count, the catalog adds Kling 3.0 Pro, Seedance 2.5, Wan 3.0, Google's Veo 3.1, xAI's Grok Imagine and Runway Gen-4 Turbo for video, Google's Nano Banana 2, OpenAI's GPT Image 2 and FLUX.2 for images, ElevenLabs' Eleven Music and Eleven v3 for audio, and Tripo H3.1 for 3D.

The prices are the vendor's. Protoface lists optimized MiniMax H3 inference from $0.01 per second, FastH3 from $0.0035 per second and LTX 2.5 Fast from $0.003 per second. Those figures come from Protoface's release and repeat on its own site, and they are a company's claim about its own service. SLOP TV has not tested them.

Here's how Protoface's pricing compares

Independent numbers are thin, because hosting tiers differ. Atlas Cloud, a competing inference platform, lists MiniMax H3 Developer from $0.02 a second, a standing 60 percent discount off the official rate it names. OpenRouter lists MiniMax's own hosted H3 from $0.13 a second. Both are the platforms' own figures as of October 9, 2026.

The $0.003 floor applies only to optimized open-weight models. Protoface's own catalog prices the proprietary ones higher: Veo 3.1 from $0.20 a second, Seedance 2.5 from $0.10 a second and Grok Imagine Video 1.5 from $0.08 a second, as listed on October 9, 2026.

Protoface launched the API and Protoface Studio, its browser interface, in August 2026, at the same starting prices for MiniMax H3 and LTX 2.5. The October update widens the catalog more than it changes the floor.

One caution for anyone counting sources: the Des Moines Register and Kitsap Sun versions of this announcement are reprints of the same EIN Presswire release, and the copy on Protoface's own newsroom page is the same text again. Reading all of them is not several sources.

For a developer, one key across 40 models removes the account and billing sprawl of stitching together separate provider integrations, and the API is pay-as-you-go with no subscription required for access. Whether Protoface's tuned open-weight models hold up against first-party hosting is the open question, and the $0.003 tier only pays off if the quality holds at that price.

Developers can get a free API key at protoface.com, with the model catalog at protoface.com/models and the API reference at docs.protoface.com.