ollama
Helm chart
Request a free trial
Contact our team to test out this Helm chart and related images for free. Please also indicate any other images you would like to evaluate.
Tag:
1
# Default values for ollama-helm.
2
# This is a YAML-formatted file.
3
# Declare variables to be passed into your templates.
4
5
# -- Number of replicas
6
replicaCount: 1
7
# Knative configuration
8
knative:
9
# -- Enable Knative integration
10
enabled: false
11
# -- Knative service container concurrency
12
containerConcurrency: 0
13
# -- Knative service timeout seconds
14
timeoutSeconds: 300
15
# -- Knative service response start timeout seconds
16
responseStartTimeoutSeconds: 300
17
# -- Knative service idle timeout seconds
18
idleTimeoutSeconds: 300
19
# -- Knative service annotations
20
annotations: {}
21
modelBootstrap:
22
# -- Time to keep completed Knative model bootstrap Jobs before cleanup. Set to null to disable TTL-based cleanup.
23
ttlSecondsAfterFinished: 300
24
# Docker image
25
image:
26
# -- Docker image registry
27
repository: cgr.dev/chainguard-private/ollama
28
# -- Docker pull policy
29
pullPolicy: IfNotPresent
30
# -- Docker image tag, overrides the image tag whose default is the chart appVersion.
31
tag: latest@sha256:18e9b8ee15ab2db7e073fec281b30b44168895b8887b638b80f5ebba925ba72b
32
# -- Docker registry secret names as an array
33
imagePullSecrets: []
34
# -- String to partially override template (will maintain the release name)
35
nameOverride: ""
36
# -- String to fully override template
37
fullnameOverride: ""
38
# -- String to fully override namespace
39
namespaceOverride: ""
40
# Ollama parameters
41
ollama:
42
# Port Ollama is listening on
43
port: 11434
44
gpu:
45
# -- Enable GPU integration
46
enabled: false
47
# -- Enable DRA GPU integration
48
# If enabled, it will use DRA instead of Device Driver Plugin and create a ResourceClaim and GpuClaimParameters
49
draEnabled: false
50
# -- DRA GPU DriverClass
51
draDriverClass: "gpu.nvidia.com"
52
# -- Existing DRA GPU ResourceClaim Template
53
draExistingClaimTemplate: ""
54
# -- GPU type: 'nvidia' or 'amd'
55
# If 'ollama.gpu.enabled', default value is nvidia
56
# If set to 'amd', this will add 'rocm' suffix to image tag if 'image.tag' is not override
57
# This is due cause AMD and CPU/CUDA are different images
58
type: 'nvidia'
59
# -- Specify the number of GPU
60
# If you use MIG section below then this parameter is ignored
61
number: 1
62
# -- only for nvidia cards; change to (example) 'nvidia.com/mig-1g.10gb' to use MIG slice
63
nvidiaResource: "nvidia.com/gpu"
64
# nvidiaResource: "nvidia.com/mig-1g.10gb" # example
65
# If you want to use more than one NVIDIA MIG you can use the following syntax (then nvidiaResource is ignored and only the configuration in the following MIG section is used)
66
67
mig:
68
# -- Enable multiple mig devices
69
# If enabled you will have to specify the mig devices
70
# If enabled is set to false this section is ignored
71
enabled: false
72
# -- Specify the mig devices and the corresponding number
73
devices: {}
74
# 1g.10gb: 1
75
# 3g.40gb: 1
76
models:
77
# -- List of models to pull at container startup
78
# The more you add, the longer the container will take to start if models are not present
79
# pull:
80
# - llama2
81
# - mistral
82
pull: []
83
# -- List of models to load in memory at container startup
84
# run:
85
# - llama2
86
# - mistral
87
run: []
88
# -- List of models to create at container startup, there are two options
89
# 1. Create a raw model
90
# 2. Load a model from configMaps, configMaps must be created before and are loaded as volume in "/models" directory.
91
# create:
92
# - name: llama3.1-ctx32768
93
# configMapRef: my-configmap
94
# configMapKeyRef: configmap-key
95
# - name: llama3.1-ctx32768
96
# template: |
97
# FROM llama3.1
98
# PARAMETER num_ctx 32768
99
create: []
100
# -- Automatically remove models present on the disk but not specified in the values file
101
clean: false
102
# -- Add insecure flag for pulling at container startup
103
insecure: false
104
# -- Override ollama-data volume mount path, default: "/root/.ollama"
105
mountPath: ""
106
# Service account
107
# ref: https://kubernetes.io/docs/tasks/configure-pod-container/configure-service-account/
108
serviceAccount:
109
# -- Specifies whether a service account should be created
110
create: true
111
# -- Automatically mount a ServiceAccount's API credentials?
112
automount: true
113
# -- Annotations to add to the service account
114
annotations: {}
115
# -- The name of the service account to use.
116
# If not set and create is true, a name is generated using the fullname template
117
name: ""
118
# -- Map of annotations to add to the pods
119
podAnnotations: {}
120
# -- Map of labels to add to the pods
121
podLabels: {}
122
# -- Pod Security Context
123
podSecurityContext: {}
124
# fsGroup: 2000
125
126
# -- Priority Class Name
127
priorityClassName: ""
128
# -- Container Security Context
129
securityContext: {}
130
# capabilities:
131
# drop:
132
# - ALL
133
# readOnlyRootFilesystem: true
134
# runAsNonRoot: true
135
# runAsUser: 1000
136
137
# -- Specify runtime class
138
runtimeClassName: ""
139
# Configure Service
140
service:
141
# -- Service type
142
type: ClusterIP
143
# -- Service port
144
port: 11434
145
# -- Service node port when service type is 'NodePort'
146
nodePort: 31434
147
# -- Load Balancer IP address
148
loadBalancerIP:
149
# -- Annotations to add to the service
150
annotations: {}
151
# -- Labels to add to the service
152
labels: {}
153
# -- IP Families for the service
154
ipFamilies: []
155
# - IPv4
156
# - IPv6
157
158
# -- IP Family Policy for the service
159
ipFamilyPolicy: ""
160
# SingleStack
161
# PreferDualStack
162
# RequireDualStack
163
# Configure Deployment
164
deployment:
165
# -- Labels to add to the deployment
166
labels: {}
167
# Configure the ingress resource that allows you to access the
168
ingress:
169
# -- Enable ingress controller resource
170
enabled: false
171
# -- IngressClass that will be used to implement the Ingress (Kubernetes 1.18+)
172
className: ""
173
# -- Additional annotations for the Ingress resource.
174
annotations: {}
175
# kubernetes.io/ingress.class: traefik
176
# kubernetes.io/ingress.class: nginx
177
# kubernetes.io/tls-acme: "true"
178
179
# The list of hostnames to be covered with this ingress record.
180
hosts:
181
- host: ollama.local
182
paths:
183
- path: /
184
pathType: Prefix
185
# -- The tls configuration for hostnames to be covered with this ingress record.
186
tls: []
187
# - secretName: chart-example-tls
188
# hosts:
189
# - chart-example.local
190
# Configure resource requests and limits
191
# ref: http://kubernetes.io/docs/user-guide/compute-resources/
192
resources:
193
# -- Pod requests
194
requests: {}
195
# Memory request
196
# memory: 4096Mi
197
198
# CPU request
199
# cpu: 2000m
200
201
# -- Pod limit
202
limits: {}
203
# Memory limit
204
# memory: 8192Mi
205
# CPU limit
206
# cpu: 4000m
207
208
# Configure extra options for liveness probe
209
# ref: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-probes/#configure-probes
210
livenessProbe:
211
# -- Enable livenessProbe
212
enabled: true
213
# -- Request path for livenessProbe
214
path: /
215
# -- Initial delay seconds for livenessProbe
216
initialDelaySeconds: 60
217
# -- Period seconds for livenessProbe
218
periodSeconds: 10
219
# -- Timeout seconds for livenessProbe
220
timeoutSeconds: 5
221
# -- Failure threshold for livenessProbe
222
failureThreshold: 6
223
# -- Success threshold for livenessProbe
224
successThreshold: 1
225
# Configure extra options for readiness probe
226
# ref: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-probes/#configure-probes
227
readinessProbe:
228
# -- Enable readinessProbe
229
enabled: true
230
# -- Request path for readinessProbe
231
path: /
232
# -- Initial delay seconds for readinessProbe
233
initialDelaySeconds: 30
234
# -- Period seconds for readinessProbe
235
periodSeconds: 5
236
# -- Timeout seconds for readinessProbe
237
timeoutSeconds: 3
238
# -- Failure threshold for readinessProbe
239
failureThreshold: 6
240
# -- Success threshold for readinessProbe
241
successThreshold: 1
242
# Configure autoscaling
243
autoscaling:
244
# -- Enable autoscaling
245
enabled: false
246
# -- Number of minimum replicas
247
minReplicas: 1
248
# -- Number of maximum replicas
249
maxReplicas: 100
250
# -- CPU usage to target replica
251
targetCPUUtilizationPercentage: 80
252
# -- targetMemoryUtilizationPercentage: 80
253
# -- Additional volumes on the output Deployment definition.
254
volumes: []
255
# -- - name: foo
256
# secret:
257
# secretName: mysecret
258
# optional: false
259
260
# -- Additional volumeMounts on the output Deployment definition.
261
volumeMounts: []
262
# -- - name: foo
263
# mountPath: "/etc/foo"
264
# readOnly: true
265
266
# -- Additional arguments on the output Deployment definition.
267
extraArgs: []
268
# -- Additional environments variables on the output Deployment definition.
269
# For extra OLLAMA env, please refer to https://github.com/ollama/ollama/blob/main/envconfig/config.go
270
extraEnv: []
271
# - name: OLLAMA_DEBUG
272
# value: "1"
273
274
# -- Additionl environment variables from external sources (like ConfigMap)
275
extraEnvFrom: []
276
# - configMapRef:
277
# name: my-env-configmap
278
279
# Enable persistence using Persistent Volume Claims
280
# ref: https://kubernetes.io/docs/concepts/storage/persistent-volumes/
281
persistentVolume:
282
# -- Enable persistence using PVC
283
enabled: false
284
# -- Ollama server data Persistent Volume access modes
285
# Must match those of existing PV or dynamic provisioner
286
# Ref: http://kubernetes.io/docs/user-guide/persistent-volumes/
287
accessModes:
288
- ReadWriteOnce
289
# -- Ollama server data Persistent Volume annotations
290
annotations: {}
291
# -- If you'd like to bring your own PVC for persisting Ollama state, pass the name of the
292
# created + ready PVC here. If set, this Chart will not create the default PVC.
293
# Requires server.persistentVolume.enabled: true
294
existingClaim: ""
295
# -- Ollama server data Persistent Volume size
296
size: 30Gi
297
# -- Ollama server data Persistent Volume Storage Class
298
# If defined, storageClassName: <storageClass>
299
# If set to "-", storageClassName: "", which disables dynamic provisioning
300
# If undefined (the default) or set to null, no storageClassName spec is
301
# set, choosing the default provisioner. (gp2 on AWS, standard on
302
# GKE, AWS & OpenStack)
303
storageClass: ""
304
# -- Ollama server data Persistent Volume Binding Mode
305
# If defined, volumeMode: <volumeMode>
306
# If empty (the default) or set to null, no volumeBindingMode spec is
307
# set, choosing the default mode.
308
volumeMode: ""
309
# -- Subdirectory of Ollama server data Persistent Volume to mount
310
# Useful if the volume's root directory is not empty
311
subPath: ""
312
# -- Pre-existing PV to attach this claim to
313
# Useful if a CSI auto-provisions a PV for you and you want to always
314
# reference the PV moving forward
315
volumeName: ""
316
# -- Node labels for pod assignment.
317
nodeSelector: {}
318
# -- Tolerations for pod assignment
319
tolerations: []
320
# -- Affinity for pod assignment
321
affinity: {}
322
# -- Lifecycle for pod assignment (override ollama.models startup pull/run)
323
lifecycle: {}
324
# How to replace existing pods
325
updateStrategy:
326
# -- Deployment strategy can be "Recreate" or "RollingUpdate". Default is Recreate
327
type: "Recreate"
328
# -- Topology Spread Constraints for pod assignment
329
topologySpreadConstraints: {}
330
# -- Wait for a grace period
331
terminationGracePeriodSeconds: 120
332
# -- Init containers to add to the pod
333
initContainers: []
334
# - name: startup-tool
335
# image: alpine:3
336
# command: [sh, -c]
337
# args:
338
# - echo init
339
340
# -- Use the host’s ipc namespace.
341
hostIPC: false
342
# -- Use the host’s pid namespace
343
hostPID: false
344
# -- Use the host's network namespace.
345
hostNetwork: false
346
# -- Extra K8s manifests to deploy
347
extraObjects: []
348
# - apiVersion: v1
349
# kind: PersistentVolume
350
# metadata:
351
# name: aws-efs
352
# data:
353
# key: "value"
354
# - apiVersion: scheduling.k8s.io/v1
355
# kind: PriorityClass
356
# metadata:
357
# name: high-priority
358
# value: 1000000
359
# globalDefault: false
360
# description: "This priority class should be used for XYZ service pods only."
361
362
# Test connection pods
363
tests:
364
enabled: true
365
# -- Labels to add to the tests
366
labels: {}
367
# -- Annotations to add to the tests
368
annotations: {}
369
podSchedulerName: ""
370
The trusted source for open source
Talk to an expertProduct
Customers
© 2026 Chainguard, Inc. All Rights Reserved.
Chainguard® and the Chainguard logo are registered trademarks of Chainguard, Inc. in the United States and/or other countries.
The other respective trademarks mentioned on this page are owned by the respective companies and use of them does not imply any affiliation or endorsement.
Chainguard® and the Chainguard logo are registered trademarks of Chainguard, Inc. in the United States and/or other countries.
The other respective trademarks mentioned on this page are owned by the respective companies and use of them does not imply any affiliation or endorsement.