LONG AI
Home › Blog › Oct 21 Model Plaza Overhaul: 174 Old Nam

Oct 21 Model Plaza Overhaul: 174 Old Names Retired, With a Migration Table

Category:资讯 · 6 min read · 2026-10-21

On the morning of Oct 21, we pulled the plaza's public API (ai.dijinlong.com/api/v1/model-plaza) and diffed the returned model names against the Oct 20 snapshot: 574 names became 593, a net gain of 19; within the list, 174 old call names disappeared and 193 new entries appeared. (Source: LONG AI model plaza, ai.dijinlong.com/model-plaza, fetched 2026-10-21 09:40; comparison snapshot 2026-10-20)

Conclusion first: this wasn't an ordinary "new arrivals" drop - it was a catalog restructuring. Of the 174 vanished names, 141 are genuinely delisted with no same-name replacement; the other 33 just moved to a new channel prefix, the model still there. The same window brought in 193 entries including Claude Opus 5.5, Kimi K3, Gemini 3.8 Flash, and MiniMax M3.1 Flash Preview.

1. What actually changed

1. 141 models truly delisted; Doubao and older Qwen took the worst hit. By name family: Doubao/Seed legacy (1.5, old pro/lite, seedream 3.0, etc.) 58, older Qwen and old snapshots 38, image/video/3D 20, embedding/OCR and others 16, Wenxin ERNIE 8, GPT 1. Doubao's doubao-pro-32k-240828 and doubao-lite-32k-240828, Wenxin's ernie-5.1 and ernie-4.5-turbo-32k, and Qwen's qwen2-7b-instruct are all gone from the plaza as of today.

2. 33 were just renames - don't mistake them for delistings. Typical: GLM 5.x bare name became ZHIPU/GLM-5.3, GPT-5.4 became cb/gpt-5.4, DeepSeek-V4-Pro became cbcn/deepseek-v4-pro, Qwen3.8-Max became ali/qwen3.8-max. The model didn't change - the call ID did.

3. 193 new entries went live. 71 of them are prefix-free new model names, including claude-opus-5.5, claude-fable-5.1, kimi-k3, gemini-3.8-flash, minimax-m3.1-flash-preview, grok-imagine-1.5, glm-5.3-prime. For users, that's a real expansion of selectable models.

2. What it means for ordinary users

Can you still use them: old names are no longer selectable on the plaza. Search for doubao-pro-32k-240828, ernie-5.1, or qwen2-7b-instruct and you get "no matching models." Renames still work, but you must switch to the full prefixed ID; true delistings have no same-name replacement and force a model change.

Did prices change: this was a list adjustment, not a repricing. The API's price fields returned unchanged; we saw no uniform hike or cut. Specific unit prices still follow the plaza page.

Should you switch: Doubao 1.5, Wenxin ERNIE, and Qwen2 users have to switch today; anyone already on Claude, GPT, or Kimi with full prefixes written in is unaffected.

LONG AI model plaza search for ernie returns 0 matches (2026-10-21)
Plaza search for "ernie": 0 matches (2026-10-21 09:45). The "467" at the top is the current number of in-stock priced models; the API's full set is 593 unique names.

3. Reproducible data points

Fetch method, for reproducibility: GET https://ai.dijinlong.com/api/v1/model-plaza, 2026-10-21 09:40, returned 2 groups, 621 records, 593 unique model names after dedup; diffed against the 2026-10-20 snapshot (574).

Per-name searches (same time, same API): doubao-pro-32k-240828 hit 0, ernie-5.1 hit 0, qwen2-7b-instruct hit 0; doubao-seed-2-0-pro-260215, claude-opus-5.5, and minimax-m3.1-flash-preview each hit 1. In other words, the old names really are gone and the new ones really are on sale.

Oct 21 plaza list change table: 174 old names gone, 141 truly delisted, 33 renamed only, 193 new entries
Oct 21 plaza changes: 174 gone, 141 delisted, 33 renamed, 193 new; see the footnote for the counting method.

4. Our advice

(Data as of 2026-10-21 09:40; the list follows the plaza's live results; this article is not a price commitment.)

相关文章

GPT-6.2 Cuts Prices 20%: What the AI Price War Means

GPT-6.2 cuts list prices 20%, yet platform discounts run deeper: gpt-6-astra trades at under 1% of list. What this week's AI price war means for buyer

What Is Prompt Caching? How Cache Tokens Cut API Bills

Prompt caching stores a repeated prompt prefix and bills cache reads at 5-12% of normal input. Real prices, hit conditions, and how to cut your API bi

All 12 Claude Models Compared: Price and Best Use

Twelve Claude versions are on sale. The newest opus-5-5 costs $8.91 a month; the oldest opus-4-6 costs $49.50. Full table and a pick for each workload

Which AI Model Writes Code Best? Speed and Price Tested

gpt-5.6-terra and deepseek-v4.1-flash win on price and reliability; gemini-3.8-flash is fastest. Effective prices, measured speed, and a three-tier pi

One API Key, Every Model: How to Integrate a Multi-Model Marketplace

Point your base URL at the platform and one key covers all 138 on-sale models across 51 vendors, in both OpenAI and Anthropic request formats. Copy-pa