vastai_sdk

Name: vastai_sdk
Author: vast-ai

206stars88forks

How to install vastai_sdk?

vast-ai/vast-cli/vastai_sdk

$npx -y skills add vast-ai/vast-cli --skill vastai_sdk

Installs into the current project.

Summary

Vast.ai Python SDK — high-level API for GPU instances, volumes, serverless endpoints, and billing.

Files · 1

View on GitHub

SKILL.md

1	# Vast.ai Python SDK (`vastai` / `vastai_sdk`)
2
3	The `vastai` package provides a Python SDK for managing GPU instances, volumes, serverless endpoints, and billing on Vast.ai. The `vastai_sdk` package is a backward-compatibility shim that re-exports `vastai`.
4
5	## Installation
6
7	```bash
8	pip install vastai
9	```
10
11	For serverless and async support:
12	```bash
13	pip install "vastai[serverless]"
14	```
15
16	## Authentication
17
18	The SDK reads the API key from `~/.vast_api_key` by default. You can also pass it explicitly:
19
20	```python
21	from vastai import VastAI
22	vast = VastAI() # reads ~/.vast_api_key
23	vast = VastAI(api_key="YOUR_API_KEY") # explicit key
24	```
25
26	Get your API key from https://console.vast.ai/manage-keys/
27
28	## Backward Compatibility
29
30	The old `vastai_sdk` import still works:
31
32	```python
33	from vastai_sdk import VastAI # equivalent to: from vastai import VastAI
34	```
35
36	## VastAI Class (High-Level SDK)
37
38	```python
39	from vastai import VastAI
40	vast = VastAI(api_key=None, server_url=None, retry=3, raw=False, quiet=False)
41	```
42
43	### Instance Management
44
45	```python
46	# List all your instances
47	instances = vast.show_instances()
48
49	# Get a single instance
50	instance = vast.show_instance(id=12345)
51
52	# Search GPU offers
53	offers = vast.search_offers(query='gpu_name=RTX_4090 num_gpus>=4 reliability>0.99')
54
55	# Create an instance from an offer
56	result = vast.create_instance(id=<offer_id>, image="pytorch/pytorch:latest", disk=50)
57
58	# Lifecycle
59	vast.start_instance(id=12345)
60	vast.stop_instance(id=12345)
61	vast.reboot_instance(id=12345)
62	vast.destroy_instance(id=12345)
63
64	# Label an instance
65	vast.label_instance(id=12345, label="my-training-run")
66
67	# Get SSH connection string
68	ssh_url = vast.ssh_url(id=12345) # returns "ssh -p PORT user@host"
69	scp_url = vast.scp_url(id=12345) # returns scp-compatible URL
70	```
71
72	### Interruptible (spot) rentals
73
74	Interruptible (spot) instances are priced below on-demand instances, but can be interrupted at any time by another user with a lower bid. Note: `vast.search_offers(type='bid', ...)` exposes `min_bid`, but `vast.create_instance(...)` defaults to on-demand at `dph_total` unless you pass `bid_price=<floor>`. Always pass `bid_price` after a `type='bid'` search, otherwise the instance will be rented as an on-demand instance/price instead of as an interruptible.
75
76	When outbid, the instance moves to `stopped` (not destroyed) and storage charges continue. Resume by raising the bid via `vast.change_bid(id=..., price=...)`.
77
78	### Search
79
80	```python
81	# Search GPU offers (use help(vast.search_offers) for full query syntax)
82	offers = vast.search_offers(query='gpu_name=RTX_3090 num_gpus>=2')
83
84	# Search volume offers
85	volumes = vast.search_volumes(query='...')
86
87	# Search network volumes
88	net_vols = vast.search_network_volumes()
89
90	# Search templates
91	templates = vast.search_templates()
92
93	# Search invoices
94	invoices = vast.search_invoices()
95	```
96
97	### Serverless Deployments
98
99	```python
100	# List all deployments
101	deployments = vast.show_deployments()
102
103	# Get a deployment
104	deployment = vast.show_deployment(id=42)
105
106	# Delete a deployment
107	vast.delete_deployment(id=42)
108	```
109
110	### Machine Management (Hosting)
111
112	```python
113	machines = vast.show_machines()
114	machine = vast.show_machine(id=10)
115	vast.list_machine(id=10, price_gpu=0.30)
116	vast.unlist_machine(id=10)
117	```
118
119	### SSH Keys
120
121	```python
122	keys = vast.show_ssh_keys()
123	vast.create_ssh_key(ssh_key="ssh-rsa AAAA...")
124	vast.delete_ssh_key(id=5)
125	```
126
127	### Team Management
128
129	```python
130	members = vast.show_members()
131	vast.invite_member(email="user@example.com", role="developer")
132	vast.remove_member(id=7)
133	```
134
135	## SyncClient (Low-Level Sync)
136
137	`SyncClient` provides typed, synchronous access to GPU offers and instances.
138
139	```python
140	from vastai import SyncClient
141
142	client = SyncClient(api_key="YOUR_API_KEY") # or reads ~/.vast_api_key
143
144	# Search offers with structured filters
145	offers = client.search(
146	num_gpus=2,
147	gpu_name="RTX_4090",
148	min_reliability=0.99,
149	max_dph_total=2.0,
150	)
151
152	# Create an instance
153	instance = client.create_instance(
154	offer_id=<id>,
155	image="pytorch/pytorch:latest",
156	disk_gb=50,
157	)
158
159	# List your instances
160	instances = client.show_instances() # returns list[SyncInstance]
161
162	# Destroy an instance
163	client.destroy_instance(instance_or_id=12345)
164	```
165
166	## AsyncClient (Low-Level Async)
167
168	`AsyncClient` provides async access to GPU offers and instances. Use as an async context manager.
169
170	```python
171	import asyncio
172	from vastai import AsyncClient
173
174	async def main():
175	async with AsyncClient(api_key="YOUR_API_KEY") as client:
176	# Search offers
177	offers = await client.search(num_gpus=1, gpu_name="A100")
178
179	# Create instance
180	instance = await client.create_instance(offer_id=<id>, image="ubuntu:22.04")
181
182	# List instances
183	instances = await client.show_instances() #

People who liked this also viewed

6 picks

Type

microsoft-foundryDeploy, evaluate, fine-tune, and manage Foundry agents end-to-end with azd: hosted agent scaffold/run/deploy, prompt agent create, batch eval, continuous eval, prompt optimizer, Agent Optimizer scaffoSkillsJul 2026476k1.3k
azure-deployExecute Azure deployments for ALREADY-PREPARED applications that have existing .azure/deployment-plan.md and infrastructure files. DO NOT use this skill when the user asks to CREATE a new application SkillsJul 2026472k1.3k
azure-preparePrepare azd-based Azure projects for deployment: generates azure.yaml, infrastructure (Bicep/Terraform), and Dockerfiles for the Azure Developer CLI (azd) workflow. USE ONLY when the user explicitly wSkillsJul 2026471k1.3k
azure-validate--- name: azure-validate description: "Pre-deployment validation for Azure readiness. Run deep checks on configuration, infrastructure (Bicep or Terraform), RBAC role assignments, managed identity pSkillsJul 2026471k1.3k
azure-aigatewayConfigure Azure API Management as an AI Gateway for AI models, MCP tools, and agents. WHEN: semantic caching, token limit, content safety, load balancing, AI model governance, MCP rate limiting, jailbSkillsJul 2026470k1.3k
azure-hosted-copilot-sdkOfficial agent plugin providing skills and MCP server configurations for Azure scenarios.SkillsJul 2026424k1.3k

vastai_sdk

206stars88forks

View on GitHub

How to install vastai_sdk?

vast-ai/vast-cli/vastai_sdk

$npx -y skills add vast-ai/vast-cli --skill vastai_sdk

Installs into the current project.

Summary

Vast.ai Python SDK — high-level API for GPU instances, volumes, serverless endpoints, and billing.

Files · 1

View on GitHub

SKILL.md

1	# Vast.ai Python SDK (`vastai` / `vastai_sdk`)
2
3	The `vastai` package provides a Python SDK for managing GPU instances, volumes, serverless endpoints, and billing on Vast.ai. The `vastai_sdk` package is a backward-compatibility shim that re-exports `vastai`.
4
5	## Installation
6
7	```bash
8	pip install vastai
9	```
10
11	For serverless and async support:
12	```bash
13	pip install "vastai[serverless]"
14	```
15
16	## Authentication
17
18	The SDK reads the API key from `~/.vast_api_key` by default. You can also pass it explicitly:
19
20	```python
21	from vastai import VastAI
22	vast = VastAI() # reads ~/.vast_api_key
23	vast = VastAI(api_key="YOUR_API_KEY") # explicit key
24	```
25
26	Get your API key from https://console.vast.ai/manage-keys/
27
28	## Backward Compatibility
29
30	The old `vastai_sdk` import still works:
31
32	```python
33	from vastai_sdk import VastAI # equivalent to: from vastai import VastAI
34	```
35
36	## VastAI Class (High-Level SDK)
37
38	```python
39	from vastai import VastAI
40	vast = VastAI(api_key=None, server_url=None, retry=3, raw=False, quiet=False)
41	```
42
43	### Instance Management
44
45	```python
46	# List all your instances
47	instances = vast.show_instances()
48
49	# Get a single instance
50	instance = vast.show_instance(id=12345)
51
52	# Search GPU offers
53	offers = vast.search_offers(query='gpu_name=RTX_4090 num_gpus>=4 reliability>0.99')
54
55	# Create an instance from an offer
56	result = vast.create_instance(id=<offer_id>, image="pytorch/pytorch:latest", disk=50)
57
58	# Lifecycle
59	vast.start_instance(id=12345)
60	vast.stop_instance(id=12345)
61	vast.reboot_instance(id=12345)
62	vast.destroy_instance(id=12345)
63
64	# Label an instance
65	vast.label_instance(id=12345, label="my-training-run")
66
67	# Get SSH connection string
68	ssh_url = vast.ssh_url(id=12345) # returns "ssh -p PORT user@host"
69	scp_url = vast.scp_url(id=12345) # returns scp-compatible URL
70	```
71
72	### Interruptible (spot) rentals
73
74	Interruptible (spot) instances are priced below on-demand instances, but can be interrupted at any time by another user with a lower bid. Note: `vast.search_offers(type='bid', ...)` exposes `min_bid`, but `vast.create_instance(...)` defaults to on-demand at `dph_total` unless you pass `bid_price=<floor>`. Always pass `bid_price` after a `type='bid'` search, otherwise the instance will be rented as an on-demand instance/price instead of as an interruptible.
75
76	When outbid, the instance moves to `stopped` (not destroyed) and storage charges continue. Resume by raising the bid via `vast.change_bid(id=..., price=...)`.
77
78	### Search
79
80	```python
81	# Search GPU offers (use help(vast.search_offers) for full query syntax)
82	offers = vast.search_offers(query='gpu_name=RTX_3090 num_gpus>=2')
83
84	# Search volume offers
85	volumes = vast.search_volumes(query='...')
86
87	# Search network volumes
88	net_vols = vast.search_network_volumes()
89
90	# Search templates
91	templates = vast.search_templates()
92
93	# Search invoices
94	invoices = vast.search_invoices()
95	```
96
97	### Serverless Deployments
98
99	```python
100	# List all deployments
101	deployments = vast.show_deployments()
102
103	# Get a deployment
104	deployment = vast.show_deployment(id=42)
105
106	# Delete a deployment
107	vast.delete_deployment(id=42)
108	```
109
110	### Machine Management (Hosting)
111
112	```python
113	machines = vast.show_machines()
114	machine = vast.show_machine(id=10)
115	vast.list_machine(id=10, price_gpu=0.30)
116	vast.unlist_machine(id=10)
117	```
118
119	### SSH Keys
120
121	```python
122	keys = vast.show_ssh_keys()
123	vast.create_ssh_key(ssh_key="ssh-rsa AAAA...")
124	vast.delete_ssh_key(id=5)
125	```
126
127	### Team Management
128
129	```python
130	members = vast.show_members()
131	vast.invite_member(email="user@example.com", role="developer")
132	vast.remove_member(id=7)
133	```
134
135	## SyncClient (Low-Level Sync)
136
137	`SyncClient` provides typed, synchronous access to GPU offers and instances.
138
139	```python
140	from vastai import SyncClient
141
142	client = SyncClient(api_key="YOUR_API_KEY") # or reads ~/.vast_api_key
143
144	# Search offers with structured filters
145	offers = client.search(
146	num_gpus=2,
147	gpu_name="RTX_4090",
148	min_reliability=0.99,
149	max_dph_total=2.0,
150	)
151
152	# Create an instance
153	instance = client.create_instance(
154	offer_id=<id>,
155	image="pytorch/pytorch:latest",
156	disk_gb=50,
157	)
158
159	# List your instances
160	instances = client.show_instances() # returns list[SyncInstance]
161
162	# Destroy an instance
163	client.destroy_instance(instance_or_id=12345)
164	```
165
166	## AsyncClient (Low-Level Async)
167
168	`AsyncClient` provides async access to GPU offers and instances. Use as an async context manager.
169
170	```python
171	import asyncio
172	from vastai import AsyncClient
173
174	async def main():
175	async with AsyncClient(api_key="YOUR_API_KEY") as client:
176	# Search offers
177	offers = await client.search(num_gpus=1, gpu_name="A100")
178
179	# Create instance
180	instance = await client.create_instance(offer_id=<id>, image="ubuntu:22.04")
181
182	# List instances
183	instances = await client.show_instances() #

People who liked this also viewed

6 picks

Type