v0.12.56

• Add zai/glm-5 and zai/glm-5-turbo models to model picker
• Aliases: glm, glm-5, glm-5-turbo

v0.12.30

chore: bump version to 0.12.30

v0.12.25

• Image generation docs — new docs/image-generation.md with API reference, curl/TypeScript/Python/OpenAI SDK examples, model pricing table, and /imagegen command reference
• Comprehensive docs refresh — architecture updated for dual-chain (Base + Solana), configuration updated with all env vars (CLAWROUTER_SOLANA_RPC_URL, CLAWROUTER_WORKER), troubleshooting updated for USDC-on-Solana funding, CHANGELOG backfilled for v0.11.14–v0.12.24

v0.12.24

• Preserve user-defined blockrun/ allowlist entries* — injectModelsConfig() no longer removes user-added blockrun/* allowlist entries on gateway restarts

v0.12.10 — /stats clear

What's New

/stats clear — Reset usage statistics

Users can now wipe all local usage logs and start fresh:

• /stats clear (or /stats reset) — deletes all usage-*.jsonl log files and confirms how many were removed
• DELETE /stats HTTP endpoint — programmatic access for the same operation, returns { cleared: true, deletedFiles: N }

These are local usage tracking files only — billing records on BlockRun's server are unaffected.

Use cases

• Clean metrics after testing
• Fresh start for a new billing period
• Benchmark a new routing config without old data

Install / Update

openclaw plugins install @blockrun/clawrouter

v0.12.9

Fix: Gemini 3 models break tool calling (#73)

• Removed toolCalling from google/gemini-3-flash-preview — this model returns empty responses on tool calls, so it's now excluded from tool-heavy routing entirely
• Added google/gemini-3-pro-preview and google/gemini-3.1-pro to TOOL_NONCOMPLIANT_MODELS — these models return tool JSON as plain text instead of structured tool_calls, so they're deprioritized when better alternatives exist in the fallback chain

v0.12.8 — GPT-5.4 & GPT-5.4 Pro

New Models

• GPT-5.4 (openai/gpt-5.4) — $2.50/$15 per 1M tokens, 400K context, reasoning + vision + agentic + tool calling
• GPT-5.4 Pro (openai/gpt-5.4-pro) — $30/$180 per 1M tokens, 400K context, reasoning + tool calling

Aliases

• gpt5 now points to openai/gpt-5.4 (was 5.2)
• New: gpt-5.4, gpt-5.4-pro

Smart Router Updates

• Auto COMPLEX: GPT-5.4 replaces GPT-5.2 as fallback (newer, same price tier)
• Premium COMPLEX: GPT-5.4 added as fallback option
• Agentic COMPLEX: GPT-5.4 replaces GPT-5.2 as fallback

GPT-5.4 Pro ($30/$180) is not in routing tiers — request it explicitly via blockrun/openai/gpt-5.4-pro.

Update

bash ~/.openclaw/extensions/clawrouter/scripts/update.sh

v0.12.5

Bug Fix

• Force agentic tiers when tools are present in API request — OpenClaw (and other agentic clients) always send tools in their API requests, but the router only checked prompt keywords to decide whether to use agentic-capable models. Simple prompts like "create a markdown file" scored too low on keyword detection and got routed to cheaper models that don't handle tool use reliably (the "model just talks instead of creating files" bug). Now when the request contains tools, agentic tiers are forced automatically — ensuring Claude, GPT, and other proven tool-calling models are used.

Thanks to @carlosdss22 for reporting!

v0.12.4

Bug Fix

• Fix Solana wallet migration missing signatures — signTransactionMessageWithSigners was producing unsigned transactions because only plain Address strings were in the instruction account metas. Added addSignersToTransactionMessage to bind both the fee payer (new wallet) and token transfer authority (old wallet) before signing. Fixes "Transaction is missing signatures for addresses" error on /wallet migrate-solana.

v0.12.3

Fixes

• [P1] Multi-account sweep: Transfer from each USDC token account individually instead of accumulating balance but only keeping the last account — fixes failure when USDC is split across multiple accounts
• [P2] Partial file reads: fs-read now loops until all bytes are consumed, respecting bytesRead return value
• [P3] JSONL resilience: Stats parser skips malformed lines instead of dropping all entries when one line fails

Full Changelog: v0.12.2...v0.12.3