docs: rebuild the free-model lineup after NVIDIA's 2026-08-30 sweep - #68
Merged
Conversation
added 4 commits
August 30, 2026 21:39
BlockRun's 402 responses now spread x402Version/accepts at the top level of the JSON body, not just the signed headers — v1-era x402 clients that only parse the body were silently failing to auto-pay. Updates every endpoint doc's 402 example plus the x402 protocol pages to match.
A live two-pass --real probe found four of the five VISIBLE free models gone at once: step-3.7-flash, nemotron-nano-9b-v2 and nemotron-nano-12b-v2-vl return a published 410, and mistral-nemotron is the quiet kind — still listed by NVIDIA, but a completion never comes back (>150s zero bytes, both passes). nemotron-super-49b (hidden, but the free cascade's tertiary rung) is 410 too. Every reference to those five is repointed at a probe-verified replacement: nemotron-3.5-lightning, nemotron-3-nano-30b, and llama-3.2-11b-vision — the only Llama NVIDIA still serves. Counts move with the catalog: 71 -> 70 chat, 95 -> 94 total, 5 -> 4 free.
Two free models (nemotron-3.5-lightning, nemotron-3-nano-omni) now serve from OpenRouter's $0 ":free" pool with the direct-NVIDIA path as fallback: same models, larger pool, 4.9s median against 16.3s, and Lightning gains a 1M context. That pool also made three more models listable — nemotron-3-ultra-550b (550B/55B, 1M ctx, unreachable on our own key), cohere/north-mini-code and poolside/laguna-xs-2.1. Counts: 70 -> 73 chat, 94 -> 97 total, 4 -> 7 free.
2 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Docs half of BlockRunAI/blockrun#447.
A live two-pass
--realprobe found four of the five VISIBLE free models gone at once —step-3.7-flash,nemotron-nano-9b-v2andnemotron-nano-12b-v2-vlreturn a published 410, andmistral-nemotronis the quiet kind: still listed by NVIDIA, but a completion never comes back (>150s zero bytes, both passes).nemotron-super-49b(hidden, but the free cascade's tertiary rung) is 410 too.Every reference to those five is repointed at a probe-verified replacement:
nvidia/nemotron-3.5-lightning— thinking-mode reasoning, 131K, ~35 tok/snvidia/nemotron-3-nano-30b— ~121 tok/s, the fastest free model in the catalognvidia/llama-3.2-11b-vision— the only Llama NVIDIA still serves; 128K, image inputCounts move with the catalog: 71 → 70 chat, 95 → 94 total, 5 → 4 free. Dated changelog entries keep their original numbers.