Skip to content

Commit f1f6b14

Browse files
canyugsclaude
canyugs
and
claude
committed
feat: sync models with Zeabur AI Hub website
Add 8 missing models: GPT-5.2, GPT-5.1, GPT-5 Nano, GLM-4.7, GLM-4.7 Flash, Kimi 2.5, Gemini 3 Pro Image Preview, and Gemini 2.5 Flash Image. Remove deepseek-v3.2-exp (not on website). Bump version to 0.0.11. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
1 parent 5b0066f commit f1f6b14

File tree

10 files changed

+487
-13
lines changed

10 files changed

+487
-13
lines changed

‎manifest.yaml‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -1,4 +1,4 @@
1-
version: 0.0.10
1+
version: 0.0.11
22
type: plugin
33
author: "zeabur"
44
name: "zeabur_ai_hub"
@@ -22,7 +22,7 @@ plugins:
2222
models:
2323
- "provider/zeabur_ai_hub.yaml"
2424
meta:
25-
version: 0.0.10
25+
version: 0.0.11
2626
minimum_dify_version: "1.0.0"
2727
arch:
2828
- amd64

‎models/llm/_position.yaml‎

Lines changed: 11 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -4,12 +4,17 @@
44
# Gemini
55
- gemini-3-pro-preview
66
- gemini-3-flash-preview
7+
- gemini-3-pro-image-preview
78
- gemini-2.5-pro
89
- gemini-2.5-flash
910
- gemini-2.5-flash-lite
11+
- gemini-2.5-flash-image
1012
# GPT
13+
- gpt-5.2
14+
- gpt-5.1
1115
- gpt-5
1216
- gpt-5-mini
17+
- gpt-5-nano
1318
- gpt-4.1
1419
- gpt-4.1-mini
1520
- gpt-4o
@@ -19,10 +24,14 @@
1924
- grok-4-fast-non-reasoning
2025
# DeepSeek
2126
- deepseek-v3.2
22-
- deepseek-v3.2-exp
23-
# Others
27+
# GLM
28+
- glm-4.7
29+
- glm-4.7-flash
2430
- glm-4.6
31+
# Kimi
2532
- kimi-k2-thinking
33+
- kimi-2.5
34+
# Others
2635
- llama-3.3-70b
2736
- qwen-3-32
2837
- qwen3-next

‎models/llm/gemini-2.5-flash-image.yaml‎

Lines changed: 67 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,67 @@
1+
model: gemini-2.5-flash-image
2+
label:
3+
en_US: Gemini 2.5 Flash Image
4+
zh_Hans: Gemini 2.5 Flash Image
5+
zh_Hant: Gemini 2.5 Flash Image
6+
model_type: llm
7+
features:
8+
- agent-thought
9+
- vision
10+
model_properties:
11+
mode: chat
12+
context_size: 1048576
13+
parameter_rules:
14+
- name: temperature
15+
use_template: temperature
16+
- name: top_p
17+
use_template: top_p
18+
- name: top_k
19+
label:
20+
en_US: Top K
21+
zh_Hans: Top K
22+
zh_Hant: Top K
23+
type: int
24+
min: 0
25+
max: 100
26+
required: false
27+
help:
28+
en_US: Limits token selection to top K choices. 0 means no limit.
29+
zh_Hans: 将令牌选择限制在前 K 个选项。0 表示无限制。
30+
zh_Hant: 將令牌選擇限制在前 K 個選項。0 表示無限制。
31+
- name: max_tokens
32+
use_template: max_tokens
33+
required: true
34+
default: 8192
35+
min: 1
36+
max: 65536
37+
- name: presence_penalty
38+
use_template: presence_penalty
39+
- name: frequency_penalty
40+
use_template: frequency_penalty
41+
- name: seed
42+
label:
43+
en_US: Seed
44+
zh_Hans: 随机种子
45+
zh_Hant: 隨機種子
46+
type: int
47+
required: false
48+
help:
49+
en_US: Random seed for reproducible outputs. Set a specific value for deterministic results.
50+
zh_Hans: 用于可重复输出的随机种子。设置特定值以获得确定性结果。
51+
zh_Hant: 用於可重複輸出的隨機種子。設置特定值以獲得確定性結果。
52+
- name: response_format
53+
label:
54+
en_US: Response Format
55+
zh_Hans: 响应格式
56+
zh_Hant: 回應格式
57+
type: string
58+
options:
59+
- text
60+
- json_object
61+
required: false
62+
63+
pricing:
64+
input: "0.3"
65+
output: "30"
66+
unit: "0.000001"
67+
currency: USD

‎models/llm/gemini-3-pro-image-preview.yaml‎

Lines changed: 67 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,67 @@
1+
model: gemini-3-pro-image-preview
2+
label:
3+
en_US: Gemini 3 Pro Image Preview
4+
zh_Hans: Gemini 3 Pro Image Preview
5+
zh_Hant: Gemini 3 Pro Image Preview
6+
model_type: llm
7+
features:
8+
- agent-thought
9+
- vision
10+
model_properties:
11+
mode: chat
12+
context_size: 1048576
13+
parameter_rules:
14+
- name: temperature
15+
use_template: temperature
16+
- name: top_p
17+
use_template: top_p
18+
- name: top_k
19+
label:
20+
en_US: Top K
21+
zh_Hans: Top K
22+
zh_Hant: Top K
23+
type: int
24+
min: 0
25+
max: 100
26+
required: false
27+
help:
28+
en_US: Limits token selection to top K choices. 0 means no limit.
29+
zh_Hans: 将令牌选择限制在前 K 个选项。0 表示无限制。
30+
zh_Hant: 將令牌選擇限制在前 K 個選項。0 表示無限制。
31+
- name: max_tokens
32+
use_template: max_tokens
33+
required: true
34+
default: 8192
35+
min: 1
36+
max: 65536
37+
- name: presence_penalty
38+
use_template: presence_penalty
39+
- name: frequency_penalty
40+
use_template: frequency_penalty
41+
- name: seed
42+
label:
43+
en_US: Seed
44+
zh_Hans: 随机种子
45+
zh_Hant: 隨機種子
46+
type: int
47+
required: false
48+
help:
49+
en_US: Random seed for reproducible outputs. Set a specific value for deterministic results.
50+
zh_Hans: 用于可重复输出的随机种子。设置特定值以获得确定性结果。
51+
zh_Hant: 用於可重複輸出的隨機種子。設置特定值以獲得確定性結果。
52+
- name: response_format
53+
label:
54+
en_US: Response Format
55+
zh_Hans: 响应格式
56+
zh_Hant: 回應格式
57+
type: string
58+
options:
59+
- text
60+
- json_object
61+
required: false
62+
63+
pricing:
64+
input: "2"
65+
output: "120"
66+
unit: "0.000001"
67+
currency: USD

‎models/llm/glm-4.7-flash.yaml‎

Lines changed: 41 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,41 @@
1+
model: glm-4.7-flash
2+
label:
3+
en_US: GLM-4.7 Flash
4+
zh_Hans: GLM-4.7 Flash
5+
zh_Hant: GLM-4.7 Flash
6+
model_type: llm
7+
features:
8+
- agent-thought
9+
- vision
10+
- tool-call
11+
- stream-tool-call
12+
model_properties:
13+
mode: chat
14+
context_size: 128000
15+
parameter_rules:
16+
- name: temperature
17+
use_template: temperature
18+
- name: top_p
19+
use_template: top_p
20+
- name: max_tokens
21+
use_template: max_tokens
22+
required: true
23+
default: 4096
24+
min: 1
25+
max: 8192
26+
- name: response_format
27+
label:
28+
en_US: Response Format
29+
zh_Hans: 响应格式
30+
zh_Hant: 回應格式
31+
type: string
32+
options:
33+
- text
34+
- json_object
35+
required: false
36+
37+
pricing:
38+
input: "0.06"
39+
output: "0.4"
40+
unit: "0.000001"
41+
currency: USD

‎models/llm/glm-4.7.yaml‎

Lines changed: 41 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,41 @@
1+
model: glm-4.7
2+
label:
3+
en_US: GLM-4.7
4+
zh_Hans: GLM-4.7
5+
zh_Hant: GLM-4.7
6+
model_type: llm
7+
features:
8+
- agent-thought
9+
- vision
10+
- tool-call
11+
- stream-tool-call
12+
model_properties:
13+
mode: chat
14+
context_size: 128000
15+
parameter_rules:
16+
- name: temperature
17+
use_template: temperature
18+
- name: top_p
19+
use_template: top_p
20+
- name: max_tokens
21+
use_template: max_tokens
22+
required: true
23+
default: 4096
24+
min: 1
25+
max: 8192
26+
- name: response_format
27+
label:
28+
en_US: Response Format
29+
zh_Hans: 响应格式
30+
zh_Hant: 回應格式
31+
type: string
32+
options:
33+
- text
34+
- json_object
35+
required: false
36+
37+
pricing:
38+
input: "0.4"
39+
output: "1.9"
40+
unit: "0.000001"
41+
currency: USD

‎models/llm/gpt-5-nano.yaml‎

Lines changed: 92 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,92 @@
1+
model: gpt-5-nano
2+
label:
3+
en_US: GPT-5 Nano
4+
zh_Hans: GPT-5 Nano
5+
zh_Hant: GPT-5 Nano
6+
model_type: llm
7+
features:
8+
- agent-thought
9+
- vision
10+
- tool-call
11+
- stream-tool-call
12+
model_properties:
13+
mode: chat
14+
context_size: 128000
15+
parameter_rules:
16+
- name: temperature
17+
use_template: temperature
18+
- name: top_p
19+
use_template: top_p
20+
- name: max_tokens
21+
use_template: max_tokens
22+
required: true
23+
default: 8192
24+
min: 1
25+
max: 32768
26+
- name: presence_penalty
27+
use_template: presence_penalty
28+
- name: frequency_penalty
29+
use_template: frequency_penalty
30+
- name: response_format
31+
label:
32+
en_US: Response Format
33+
zh_Hans: 响应格式
34+
zh_Hant: 回應格式
35+
type: string
36+
options:
37+
- text
38+
- json_object
39+
required: false
40+
- name: json_schema
41+
label:
42+
en_US: JSON Schema
43+
zh_Hans: JSON Schema
44+
zh_Hant: JSON Schema
45+
type: text
46+
required: false
47+
help:
48+
en_US: Define a JSON Schema for structured output. Only works when Response Format is json_object.
49+
zh_Hans: 定义 JSON Schema 以获取结构化输出。仅在响应格式为 json_object 时有效。
50+
zh_Hant: 定義 JSON Schema 以獲取結構化輸出。僅在回應格式為 json_object 時有效。
51+
- name: seed
52+
label:
53+
en_US: Seed
54+
zh_Hans: 随机种子
55+
zh_Hant: 隨機種子
56+
type: int
57+
required: false
58+
help:
59+
en_US: Random seed for deterministic sampling. Use with system_fingerprint for reproducibility.
60+
zh_Hans: 用于确定性采样的随机种子。与 system_fingerprint 一起使用以实现可重复性。
61+
zh_Hant: 用於確定性採樣的隨機種子。與 system_fingerprint 一起使用以實現可重複性。
62+
- name: logprobs
63+
label:
64+
en_US: Log Probabilities
65+
zh_Hans: 对数概率
66+
zh_Hant: 對數機率
67+
type: boolean
68+
default: false
69+
required: false
70+
help:
71+
en_US: Whether to return log probabilities of the output tokens.
72+
zh_Hans: 是否返回输出令牌的对数概率。
73+
zh_Hant: 是否返回輸出令牌的對數機率。
74+
- name: top_logprobs
75+
label:
76+
en_US: Top Log Probabilities
77+
zh_Hans: 顶部对数概率
78+
zh_Hant: 頂部對數機率
79+
type: int
80+
min: 0
81+
max: 20
82+
required: false
83+
help:
84+
en_US: Number of most likely tokens to return at each position (0-20). Requires logprobs to be true.
85+
zh_Hans: 每个位置返回的最可能令牌数量 (0-20)。需要 logprobs 为 true。
86+
zh_Hant: 每個位置返回的最可能令牌數量 (0-20)。需要 logprobs 為 true。
87+
88+
pricing:
89+
input: "0.05"
90+
output: "0.4"
91+
unit: "0.000001"
92+
currency: USD

0 commit comments

Comments
 (0)