twoimo · twoimo · Jul 25, 2026 · Jul 25, 2026
diff --git a/docs/gpt-5.6-codex-preset-benchmark.md b/docs/gpt-5.6-codex-preset-benchmark.md
@@ -9,7 +9,7 @@ Built-in role assignments are product judgments. The selected TypeScript edit ev
 - **Eco**: `terra:low` default, `luna:low` executor, `luna:high` planner, `terra:xhigh` critic, and `terra:high` architect.
 - **Medium**: `sol:low` default, `terra:low` executor, `terra:high` planner, `sol:xhigh` critic, and `sol:high` architect.
 - **Pro**: `sol:medium` default, `terra:medium` executor, `sol:high` planner, `sol:max` critic, and `sol:xhigh` architect.
-- **Combos**: `opus-codex` uses the Medium Codex executor, critic, and architect roles, with the durable `anthropic/claude-sonnet-5` planner override; `codex-opencodego` uses Medium Codex default and architect roles; and `fable-opus-codex` uses Pro Codex executor and architect roles with `anthropic/claude-opus-4-8:medium` as planner.
+- **Combos**: `opus-codex` uses the Medium Codex executor, critic, and architect roles, with the durable `anthropic/claude-sonnet-5` planner override; `codex-opencodego` uses Medium Codex default and architect roles; and `fable-opus-codex` uses Pro Codex executor and architect roles with `anthropic/claude-opus-5:medium` as planner.
 
 The edit benchmark does not measure default-agent interpretation, orchestration, explanation, or routing, and it does not measure planner, architect, or critic work. Those non-executor assignments are product judgments, not benchmark findings.
 
@@ -119,9 +119,9 @@ The selected-task data show that Luna xhigh used more reported tokens than Luna
 | `codex-eco` | `openai-codex/gpt-5.6-terra:low` | `openai-codex/gpt-5.6-luna:low` | `openai-codex/gpt-5.6-luna:high` | `openai-codex/gpt-5.6-terra:xhigh` | `openai-codex/gpt-5.6-terra:high` |
 | `codex-medium` | `openai-codex/gpt-5.6-sol:low` | `openai-codex/gpt-5.6-terra:low` | `openai-codex/gpt-5.6-terra:high` | `openai-codex/gpt-5.6-sol:xhigh` | `openai-codex/gpt-5.6-sol:high` |
 | `codex-pro` | `openai-codex/gpt-5.6-sol:medium` | `openai-codex/gpt-5.6-terra:medium` | `openai-codex/gpt-5.6-sol:high` | `openai-codex/gpt-5.6-sol:max` | `openai-codex/gpt-5.6-sol:xhigh` |
-| `opus-codex` | `anthropic/claude-opus-4-8:xhigh` | `openai-codex/gpt-5.6-terra:low` | `anthropic/claude-sonnet-5` | `openai-codex/gpt-5.6-sol:xhigh` | `openai-codex/gpt-5.6-sol:high` |
+| `opus-codex` | `anthropic/claude-opus-5:xhigh` | `openai-codex/gpt-5.6-terra:low` | `anthropic/claude-sonnet-5` | `openai-codex/gpt-5.6-sol:xhigh` | `openai-codex/gpt-5.6-sol:high` |
 | `codex-opencodego` | `openai-codex/gpt-5.6-sol:low` | `opencode-go/deepseek-v4-pro` | `opencode-go/kimi-k2.6` | `opencode-go/mimo-v2.5-pro` | `openai-codex/gpt-5.6-sol:high` |
-| `fable-opus-codex` | `anthropic/claude-fable-5:high` | `openai-codex/gpt-5.6-terra:medium` | `anthropic/claude-opus-4-8:medium` | `anthropic/claude-opus-4-8:high` | `openai-codex/gpt-5.6-sol:xhigh` |
+| `fable-opus-codex` | `anthropic/claude-fable-5:high` | `openai-codex/gpt-5.6-terra:medium` | `anthropic/claude-opus-5:medium` | `anthropic/claude-opus-5:high` | `openai-codex/gpt-5.6-sol:xhigh` |
 
 ## Limitations
 

diff --git a/docs/models.md b/docs/models.md
@@ -231,11 +231,11 @@ Built-in profiles are grouped by provider mix and tier:
 
 - `codex-{eco,medium,pro}` — GPT-5.6 Sol/Terra/Luna role mixes tuned by tier and reasoning effort
 - `opencodego` — single OpenCode Go preset (Kimi default, DeepSeek executor/architect, Qwen planner, MiMo critic)
-- `claude-opus` — Anthropic OAuth preset centered on `claude-opus-4-8`
+- `claude-opus` — Anthropic OAuth preset centered on `claude-opus-5`
 - Single-provider tiers: `glm-{eco,medium,pro}`, `kimi-coding-plan-{eco,medium,pro}`, `mimo-{eco,medium,pro}`, `grok-{eco,medium,pro}`, `cursor-{eco,medium,pro}`, `minimax-{eco,medium,pro}`
 - Combos: `opus-codex`, `codex-opencodego`, and `fable-opus-codex`
 
-The `eco`, `medium`, and `pro` Codex profile mappings are current product judgments: Eco assigns Terra low/Luna low/Luna high/Terra xhigh/Terra high to default/executor/planner/critic/architect; Medium assigns Sol low/Terra low/Terra high/Sol xhigh/Sol high; and Pro assigns Sol medium/Terra medium/Sol high/Sol max/Sol xhigh. `opus-codex` retains the Medium Codex executor, critic, and architect roles but uses `anthropic/claude-sonnet-5` for planner; `codex-opencodego` retains the Medium Codex default and architect roles; and `fable-opus-codex` uses the Pro Codex executor and architect roles with `anthropic/claude-opus-4-8:medium` for planner. The descriptive repeated local exact-edit evidence informs only selected executor-style TypeScript tasks; it does not evaluate or prove default, planner, architect, or critic performance. See [GPT-5.6 Codex preset benchmark](./gpt-5.6-codex-preset-benchmark.md). Effort suffixes are clamped to each model's supported thinking range at preview and activation time. Single-provider tiers pin each provider's current flagship (`zai/glm-5.2`, `kimi-code/kimi-k2.7-code`, `xiaomi/mimo-v2.5-pro`, `xai/grok-4.3`, `cursor/composer-1.5`, `minimax-code/minimax-m3`). User-defined profiles override built-ins by exact profile name.
+The `eco`, `medium`, and `pro` Codex profile mappings are current product judgments: Eco assigns Terra low/Luna low/Luna high/Terra xhigh/Terra high to default/executor/planner/critic/architect; Medium assigns Sol low/Terra low/Terra high/Sol xhigh/Sol high; and Pro assigns Sol medium/Terra medium/Sol high/Sol max/Sol xhigh. `opus-codex` retains the Medium Codex executor, critic, and architect roles but uses `anthropic/claude-sonnet-5` for planner; `codex-opencodego` retains the Medium Codex default and architect roles; and `fable-opus-codex` uses the Pro Codex executor and architect roles with `anthropic/claude-opus-5:medium` for planner. The descriptive repeated local exact-edit evidence informs only selected executor-style TypeScript tasks; it does not evaluate or prove default, planner, architect, or critic performance. See [GPT-5.6 Codex preset benchmark](./gpt-5.6-codex-preset-benchmark.md). Effort suffixes are clamped to each model's supported thinking range at preview and activation time. Single-provider tiers pin each provider's current flagship (`zai/glm-5.2`, `kimi-code/kimi-k2.7-code`, `xiaomi/mimo-v2.5-pro`, `xai/grok-4.3`, `cursor/composer-1.5`, `minimax-code/minimax-m3`). User-defined profiles override built-ins by exact profile name.
 
 
 Use `gjc --mpreset <name>` to activate a profile for the current session only. Activation hard-blocks when any provider listed in `required_providers` lacks credentials. Add `--default` to persist the selected profile as `modelProfile.default` in `config.yml`, so it applies at startup:

diff --git a/docs/multi-vendor-profiles.md b/docs/multi-vendor-profiles.md
@@ -2,7 +2,7 @@
 
 A practical guide to picking models for GJC's roles, for every subscription situation — one vendor, two vendors, or the full multi-vendor set. It adds curated cross-vendor `profiles:` for `~/.gjc/agent/models.yml` and verified selector notes on top of the mechanism in [Model profiles](./models.md#model-profiles---mpreset). Everything here is **user config**; it complements the built-in `--mpreset` presets and overrides a built-in only when it shares its exact name.
 
-> Selectors, prices, and "axis leaders" are catalog- and time-sensitive (observed 2026-06 on the current bundled catalog). Re-verify any selector with `gjc -p --no-session --no-tools --model <selector> "Reply OK"`.
+> Selectors, prices, and "axis leaders" are catalog- and time-sensitive (selectors and prices observed 2026-07 on the current bundled catalog; the measured latency and single-message-limit notes below were observed 2026-06 on `claude-opus-4-8` and have not been re-measured on `claude-opus-5`). Re-verify any selector with `gjc -p --no-session --no-tools --model <selector> "Reply OK"`.
 
 ## The five roles
 
@@ -36,7 +36,7 @@ profiles:
   daily:                 # everyday balance
     required_providers: [anthropic, openai-codex, google-antigravity, xai]
     model_mapping:
-      default:   anthropic/claude-opus-4-8:medium
+      default:   anthropic/claude-opus-5:medium
       executor:  openai-codex/gpt-5.4:high
       planner:   google-antigravity/gemini-3.1-pro-low:high
       architect: google-antigravity/gemini-3.1-pro-low:high
@@ -45,16 +45,16 @@ profiles:
   ultimate:              # cost-no-object, best per role
     required_providers: [anthropic, openai-codex, google-antigravity, xai]
     model_mapping:
-      default:   anthropic/claude-opus-4-8:high
-      executor:  anthropic/claude-opus-4-8:max
+      default:   anthropic/claude-opus-5:high
+      executor:  anthropic/claude-opus-5:max
       planner:   openai-codex/gpt-5.5:xhigh
       architect: google-antigravity/gemini-3.1-pro-low:high
       critic:    xai/grok-4.3:high
 
   eco:                   # cheapest delegated work; main loop stays on Opus
     required_providers: [anthropic, opencode-go, google-antigravity, xai]
     model_mapping:
-      default:   anthropic/claude-opus-4-8:low
+      default:   anthropic/claude-opus-5:low
       executor:  opencode-go/deepseek-v4-flash
       planner:   xai/grok-4-1-fast:high
       architect: google-antigravity/gemini-3.1-pro-low
@@ -63,19 +63,19 @@ profiles:
   monorepo:              # huge codebases (openai-codex excluded: 272k context cap)
     required_providers: [anthropic, google-antigravity, opencode-go]
     model_mapping:
-      default:   anthropic/claude-opus-4-8:medium
-      executor:  anthropic/claude-opus-4-8:high
+      default:   anthropic/claude-opus-5:medium
+      executor:  anthropic/claude-opus-5:high
       planner:   google-antigravity/gemini-3.1-pro-low:high
-      architect: anthropic/claude-opus-4-8:high
+      architect: anthropic/claude-opus-5:high
       critic:    opencode-go/glm-5.2
 
   reviewer:              # review/audit stance — the author-mode role split, inverted
     required_providers: [anthropic, openai-codex, google-antigravity]
     model_mapping:
-      default:   anthropic/claude-opus-4-8:high                 # aggregator restraint: preserve raw reviewer verdicts
+      default:   anthropic/claude-opus-5:high                 # aggregator restraint: preserve raw reviewer verdicts
       executor:  openai-codex/gpt-5.5:high                      # support — repro PoCs, failing tests, harnesses
       planner:   google-antigravity/gemini-3.1-pro-low:high     # review checklists / audit scoping
-      architect: anthropic/claude-opus-4-8:high                 # lead 1 — primary code-review judge (effective long-context)
+      architect: anthropic/claude-opus-5:high                 # lead 1 — primary code-review judge (effective long-context)
       critic:    openai-codex/gpt-5.5:high                      # lead 2 — merge gate, cross-family vs Claude-authored code
 ```
 
@@ -101,22 +101,22 @@ Current axis leaders and the cheaper second option, with metered price ($/1M in/
 
 | Need | First pick | Cheaper option |
 | --- | --- | --- |
-| Router / tool-calling (`default`) | `anthropic/claude-opus-4-8` (5/25) | `anthropic/claude-sonnet-5` (3/15) |
-| Coding (`executor`) | `anthropic/claude-opus-4-8` — SWE-bench Verified ~88.6 (5/25) | `openai-codex/gpt-5.4` (2.5/15) · `opencode-go/deepseek-v4-flash` (0.14/0.28) |
+| Router / tool-calling (`default`) | `anthropic/claude-opus-5` (5/25) | `anthropic/claude-sonnet-5` (3/15) |
+| Coding (`executor`) | `anthropic/claude-opus-5` (5/25) — the prior `claude-opus-4-8` scored SWE-bench Verified ~88.6; no Opus 5 measurement yet | `openai-codex/gpt-5.4` (2.5/15) · `opencode-go/deepseek-v4-flash` (0.14/0.28) |
 | Reasoning (`planner`) | `openai-codex/gpt-5.5` (ARC-AGI-2) / `google-antigravity/gemini-3.1-pro-low:high` (GPQA) | `xai/grok-4-1-fast` (0.2/0.5) |
-| Large context (`architect`) | `anthropic/claude-opus-4-8` (effective long-context) | `xai/grok-4-fast` (2M nominal, 0.2/0.5) |
+| Large context (`architect`) | `anthropic/claude-opus-5` (effective long-context) | `xai/grok-4-fast` (2M nominal, 0.2/0.5) |
 | Multimodal review (`architect`) | `google-antigravity/gemini-3.1-pro-low:high` | `google-antigravity/gemini-3.5-flash` |
 | Independent critic | `xai/grok-4.3` (1.25/2.5) | `opencode-go/glm-5.2` · `google-antigravity/gemini-3.5-flash` |
 
-On standard tasks, all current frontier models in the catalog are accurate; **pick by cost, latency, and role fit, not by raw accuracy on easy prompts.** As an indicative GJC-routed latency reference (`gjc -p`, identical coding + reasoning prompts, all correct): `grok-4.3` and `glm-5.2` ≈ 2–3s, `deepseek-v4-pro` ≈ 3–4s, `claude-opus-4-8` / `gpt-5.5` ≈ 4–7s, `gemini-3.1-pro-low:high` ≈ 7s.
+On standard tasks, all current frontier models in the catalog are accurate; **pick by cost, latency, and role fit, not by raw accuracy on easy prompts.** As an indicative GJC-routed latency reference (`gjc -p`, identical coding + reasoning prompts, all correct): `grok-4.3` and `glm-5.2` ≈ 2–3s, `deepseek-v4-pro` ≈ 3–4s, `claude-opus-4-8` / `gpt-5.5` ≈ 4–7s, `gemini-3.1-pro-low:high` ≈ 7s. `claude-opus-5` shares Opus 4.8's published context/output envelope but has not been latency-measured here.
 
 ## Verified selector notes (current catalog)
 
 Observed via live `gjc -p` calls; useful when wiring the profiles above:
 
 - **Antigravity Gemini, high reasoning** → use `google-antigravity/gemini-3.1-pro-low:high`. The id `gemini-3.1-pro-high` returns HTTP 400 (no matching backend model); `thinkingLevel` is a per-request parameter, so raising it on `gemini-3.1-pro-low` invokes the model's native high-reasoning mode rather than a degraded one.
 - **openai-codex on a ChatGPT account** serves base GPT only (`gpt-5.5`, `gpt-5.4`). Standalone `-codex` variants (`gpt-5.3-codex`, `gpt-5.2-codex`, `gpt-5.1-codex-max` / `-mini`) return `not supported when using Codex with a ChatGPT account`.
-- **Single-message input limit is separate from the context window.** `claude-opus-4-8` runs with a 1M window via multi-turn accumulation, but a single `@file` message above ~400k tokens returns 400 on `anthropic` / `google-antigravity`; `xai` / `opencode-go` accept larger single messages. Chunk very large inputs across turns instead of pasting one block.
+- **Single-message input limit is separate from the context window.** Measured on `claude-opus-4-8` (not yet re-measured on `claude-opus-5`, which publishes the same 1M window): the model runs with a 1M window via multi-turn accumulation, but a single `@file` message above ~400k tokens returns 400 on `anthropic` / `google-antigravity`; `xai` / `opencode-go` accept larger single messages. Chunk very large inputs across turns instead of pasting one block.
 - **Some selectors come from a provider's live catalog, not the bundled snapshot.** `opencode-go/glm-5.2` and `google-antigravity/gemini-3.5-flash` resolved in `gjc -p` tests but are **not** in `packages/ai/src/models.json`; they appear only after the provider's online model discovery has populated the registry. `required_providers` verifies credentials at activation — it does **not** guarantee fresh, non-stale discovery — so activation can still fail with `selector did not resolve` until discovery runs (re-login or retry to refresh). If you hit that, substitute a bundled id: `opencode-go/deepseek-v4-pro` for the critic, or `zai/glm-5.2` (add `zai` to `required_providers`) for GLM 5.2.
 
 ## Activation

diff --git a/packages/ai/scripts/generate-models.ts b/packages/ai/scripts/generate-models.ts
@@ -281,14 +281,54 @@ function applyCodexPricingFallback(models: readonly Model[]): Model[] {
 	});
 }
 
-// Catalog sources occasionally omit image input for Claude Opus 4.8 variants
+// Catalog sources occasionally omit image input for recent Claude Opus variants
 // (e.g. kilo/venice "-fast" entries) even though every Claude Opus model is
 // vision-capable. Correct those so capability advertising stays consistent
 // across providers. Runs after the dynamic merge so it survives regeneration.
+//
+// The list is an explicit allowlist of reviewed generations rather than a
+// `claude-opus-*` prefix match: a future generation must be reviewed before we
+// assert capabilities for it. `claude-opus-vision.test.ts` imports this list and
+// fails when the catalog bundles a newer Opus generation than any declared here.
+export const VISION_CORRECTED_CLAUDE_OPUS_GENERATIONS: readonly number[] = [4.8, 5];
+
+/**
+ * Known separator-less generation aliases. Upstream normally writes
+ * `claude-opus-4-5`, but a few catalogs collapse it to `claude-opus-45`. This is
+ * an explicit list so a future two-digit major (`claude-opus-10`) is read as
+ * generation 10 rather than silently as 1.0.
+ */
+const COMPACT_CLAUDE_OPUS_ALIASES: Readonly<Record<string, number>> = {
+	"41": 4.1,
+	"45": 4.5,
+	"46": 4.6,
+	"47": 4.7,
+	"48": 4.8,
+};
+
+/**
+ * Extract the Claude Opus generation from a model id, ignoring provider
+ * prefixes, region prefixes, and trailing aliases or date suffixes:
+ * `claude-opus-4-8` and `anthropic.claude-opus-4-8` -> 4.8, `claude-opus-5-fast`
+ * -> 5, `claude-opus-45` -> 4.5, `claude-opus-4-20250514` -> 4,
+ * `claude-opus-10` -> 10. Returns undefined when the id is not a Claude Opus
+ * model.
+ */
+export function claudeOpusGeneration(modelId: string): number | undefined {
+	const match = modelId
+		.toLowerCase()
+		.replace(/\./g, "-")
+		.match(/claude-opus-(\d+)(?:-(\d)(?![\d]))?/);
+	if (!match) return undefined;
+	const [, major, minor] = match;
+	if (minor !== undefined) return Number(major) + Number(minor) / 10;
+	return COMPACT_CLAUDE_OPUS_ALIASES[major] ?? Number(major);
+}
+
 function applyClaudeOpusVisionCorrections(models: readonly Model[]): Model[] {
 	return models.map(model => {
-		const normalizedId = model.id.toLowerCase().replace(/\./g, "-");
-		if (!normalizedId.includes("claude-opus-4-8")) {
+		const generation = claudeOpusGeneration(model.id);
+		if (generation === undefined || !VISION_CORRECTED_CLAUDE_OPUS_GENERATIONS.includes(generation)) {
 			return model;
 		}
 		if (model.input.includes("image")) {